EOSC510_FinalProject.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/b0ws3r/EOSC510_Final/blob/main/EOSC510_FinalProject.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "jxHmuUtdq50r",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3c0fced7-0d86-4d95-9a14-271af0d712ff"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/drive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "import csv\n",
        "import uuid\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import pickle\n",
        "import matplotlib\n",
        "from sklearn.decomposition import PCA\n",
        "from sklearn.neural_network import MLPRegressor\n",
        "from pandas import DataFrame\n",
        "import datetime\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oyBZngpcBFpT"
      },
      "source": [
        "## Definitions"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DQGnghhm6cxb"
      },
      "outputs": [],
      "source": [
        "\n",
        "wdir = '/content/drive/MyDrive/Colab Notebooks/EOSC510_Final'\n",
        "def rmse(target,prediction):\n",
        "    return(np.sqrt(((target - prediction)**2).sum()/len(target)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8tXgh4rELM4b"
      },
      "outputs": [],
      "source": [
        "def filter_on_date(df, year, month, day):\n",
        "      df_filtered = df\n",
        "      if \"Date\" in df.columns:\n",
        "        df['Date'] = pd.to_datetime(df['Date'])\n",
        "        date_filter = df[\"Date\"] == pd.Timestamp(year, month, day)\n",
        "        df_filtered = df[date_filter]\n",
        "      else:\n",
        "        print(\"No date column available to filter on\")\n",
        "      return df_filtered"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "i2dI18z5NE0x"
      },
      "outputs": [],
      "source": [
        "# we get output y_hats for 1 date. need to compare them to orig\n",
        "def join_data_to_original(predictions, original_all_columns, date_filter=None, ):\n",
        "  df_new = original_all_columns.copy()\n",
        "\n",
        "  print(predictions.columns)\n",
        "  ind = predictions.columns.shape[0]\n",
        "  print(ind)\n",
        "  df_new[\"pred_tmin\"] = predictions[\"pred_tmin\"].iloc[:].values\n",
        "  df_new[\"pred_tmax\"] = predictions[\"pred_tmax\"].iloc[:].values\n",
        "\n",
        "  print(df_new.head(3))\n",
        "  return df_new"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kO-kmSeibnzM"
      },
      "outputs": [],
      "source": [
        "def get_predictions_from_best_performer(input_data):\n",
        "  best_performer = \"/content/drive/MyDrive/Colab Notebooks/EOSC510_Final/2b1fe551-20b7-45c2-911e-338208fcdd15.p\"\n",
        "  loaded_models = pickle.load(open(best_performer,\"rb\"))\n",
        "  print(best_performer)\n",
        "  all_tmins =[]\n",
        "  all_tmaxs = []\n",
        "  for mod in loaded_models:\n",
        "    pred_output = mod.predict(input_data)\n",
        "    output_rows = pred_output.shape[0]\n",
        "    tmins = pred_output[:,0]\n",
        "    tmaxs = pred_output[:,1]\n",
        "    all_tmins.append(tmins)\n",
        "    all_tmaxs.append(tmaxs)\n",
        "\n",
        "  # get ensemble mean for each variable\n",
        "  all_tmins = np.asarray(all_tmins)\n",
        "  tmin_mean_predicted_output = all_tmins.mean(axis=0)\n",
        "  all_tmaxs = np.asarray(all_tmaxs)\n",
        "  tmax_mean_predicted_output = all_tmaxs.mean(axis=0)\n",
        "\n",
        "  data = {\"pred_tmin\": tmin_mean_predicted_output, \"pred_tmax\": tmax_mean_predicted_output}\n",
        "  predictions_df = pd.DataFrame(data)\n",
        "  print(predictions_df.head(3))\n",
        "  return predictions_df\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pGvCxhYgcnmC"
      },
      "source": [
        "## Data prep"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "CUeZde5U7n_4",
        "outputId": "a5569e1b-59d5-4471-f3a4-4423eb351175"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-6-a7653b24b4a4>:28: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
            "  val = val[val_filter_S]\n",
            "<ipython-input-6-a7653b24b4a4>:38: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
            "  y_val = y_val[date_filter]\n",
            "<ipython-input-6-a7653b24b4a4>:55: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
            "  y_tst = y_tst[date_filter]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "       Latitude  Longitude        us  gfs_tmin  gfs_tmax\n",
            "19452     48.46    -122.00 -0.804903  0.002023 -0.491285\n",
            "19453     48.46    -121.96 -0.030734  0.002023 -0.491285\n",
            "19454     48.46    -121.92  0.070596  0.002023 -0.491285\n",
            "         tmin_K    tmax_K\n",
            "19452 -0.452919 -0.574183\n",
            "19453 -0.858101 -0.952229\n",
            "19454 -0.914989 -0.998968\n",
            "      Latitude  Longitude        us  gfs_tmin  gfs_tmax\n",
            "6608     48.71    -122.00  0.903537  0.495339   -0.1041\n",
            "6609     48.71    -121.96  0.536719  0.495339   -0.1041\n",
            "6610     48.71    -121.92  0.194220  0.495339   -0.1041\n",
            "        tmin_K    tmax_K\n",
            "6608 -0.989336 -0.924240\n",
            "6609 -0.772210 -0.559578\n",
            "6610 -0.610965 -0.219740\n",
            "     Latitude  Longitude        us  gfs_tmin  gfs_tmax\n",
            "209     48.96    -122.00 -0.144225  0.685951   0.29605\n",
            "210     48.96    -121.96  0.206380  0.685951   0.29605\n",
            "211     48.96    -121.92 -0.048974  0.685951   0.29605\n",
            "       tmin_K    tmax_K\n",
            "209 -0.325715 -0.297217\n",
            "210 -0.502490 -0.504500\n",
            "211 -0.438111 -0.263990\n",
            "     Latitude  Longitude        us  gfs_tmin  gfs_tmax\n",
            "209     48.96    -122.00 -0.144225  0.685951   0.29605\n",
            "210     48.96    -121.96  0.206380  0.685951   0.29605\n",
            "211     48.96    -121.92 -0.048974  0.685951   0.29605\n",
            "       tmin_K    tmax_K\n",
            "209 -0.325715 -0.297217\n",
            "210 -0.502490 -0.504500\n",
            "211 -0.438111 -0.263990\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-6-a7653b24b4a4>:71: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
            "  y_all = all[date_filter]\n",
            "<ipython-input-6-a7653b24b4a4>:82: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  x_train[var] = (x_train[var] - x_mean) / x_std\n",
            "<ipython-input-6-a7653b24b4a4>:83: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  x_val[var] = (x_val[var] - x_mean) / x_std\n",
            "<ipython-input-6-a7653b24b4a4>:84: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  x_tst[var] = (x_tst[var] - x_mean) / x_std\n",
            "<ipython-input-6-a7653b24b4a4>:85: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  x_all[var] = (x_all[var] - x_mean) / x_std\n",
            "<ipython-input-6-a7653b24b4a4>:92: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  y_train[var] = (y_train[var] - y_mean) / y_std\n",
            "<ipython-input-6-a7653b24b4a4>:95: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  y_all[var] = (y_all[var] - y_mean) / y_std\n"
          ]
        }
      ],
      "source": [
        "### Data\n",
        "df = pd.read_pickle(f\"{wdir}/prism_gfs_all_columns_result.p\")\n",
        "\n",
        "# X  = df.iloc[:, 0:2]\n",
        "# Y = df['y']\n",
        "\n",
        "\n",
        "# training is data pts b/w 48 and 48.5 deg north\n",
        "# split X into training and validation datasets (50% train, 25% validation, 25% test)\n",
        "trn_filter = df[\"Latitude\"].astype(float) < (48.5)\n",
        "\n",
        "\n",
        "val_filter_S = df[\"Latitude\"] >= 48.5\n",
        "val_filter_N = df[\"Latitude\"] < 48.75\n",
        "\n",
        "tst_filter_S = df[\"Latitude\"] >= 48.75\n",
        "tst_filter_N = df[\"Latitude\"] <= 49\n",
        "\n",
        "# set training\n",
        "train = df[trn_filter]\n",
        "# save training data\n",
        "pickle.dump(train,open(f\"{wdir}/train_all_columns.p\",\"wb\"))\n",
        "x_train = train[[\"Latitude\", \"Longitude\",\"us\", \"gfs_tmin\", \"gfs_tmax\"]]\n",
        "y_train = train[[\"tmin_K\", \"tmax_K\"]]\n",
        "\n",
        "# validation is what we don't pick for training\n",
        "val = df[val_filter_N]\n",
        "val = val[val_filter_S]\n",
        "validation_date =pd.Timestamp(2023, 4, 15) # because why not\n",
        "val['Date'] = pd.to_datetime(val['Date'])\n",
        "date_filter = val[\"Date\"] == validation_date\n",
        "val = val[date_filter]\n",
        "pickle.dump(val,open(f\"{wdir}/val_all_columns.p\",\"wb\"))\n",
        "\n",
        "# get rid of date col\n",
        "x_val = val[[\"Latitude\", \"Longitude\",\"us\", \"gfs_tmin\", \"gfs_tmax\"]]\n",
        "y_val = val[[\"tmin_K\", \"tmax_K\", \"Date\"]]\n",
        "y_val = y_val[date_filter]\n",
        "y_val = y_val[[\"tmin_K\", \"tmax_K\", ]]\n",
        "\n",
        "\n",
        "# testing ---> don't touch until very end!!! OR ELSE\n",
        "# tst = df[tst_filter]\n",
        "test = df[tst_filter_N]\n",
        "test = test[tst_filter_S]\n",
        "validation_date =pd.Timestamp(2023, 4, 15) # because why not\n",
        "test['Date'] = pd.to_datetime(test['Date'])\n",
        "date_filter = test[\"Date\"] == validation_date\n",
        "test = test[date_filter]\n",
        "pickle.dump(test,open(f\"{wdir}/tst_all_columns.p\",\"wb\"))\n",
        "\n",
        "# get rid of date col\n",
        "x_tst = test[[\"Latitude\", \"Longitude\",\"us\", \"gfs_tmin\", \"gfs_tmax\"]]\n",
        "y_tst = test[[\"tmin_K\", \"tmax_K\", \"Date\"]]\n",
        "y_tst = y_tst[date_filter]\n",
        "y_tst = y_tst[[\"tmin_K\", \"tmax_K\", ]]\n",
        "\n",
        "\n",
        "## all data for visual\n",
        "\n",
        "all = df\n",
        "validation_date =pd.Timestamp(2023, 4, 15) # because why not\n",
        "all['Date'] = pd.to_datetime(all['Date'])\n",
        "date_filter = all[\"Date\"] == validation_date\n",
        "all = all[date_filter]\n",
        "pickle.dump(all,open(f\"{wdir}/all_all_columns.p\",\"wb\"))\n",
        "\n",
        "# get rid of date col\n",
        "x_all = all[[\"Latitude\", \"Longitude\",\"us\", \"gfs_tmin\", \"gfs_tmax\"]]\n",
        "y_all = all[[\"tmin_K\", \"tmax_K\", \"Date\"]]\n",
        "y_all = all[date_filter]\n",
        "y_all = all[[\"tmin_K\", \"tmax_K\", ]]\n",
        "\n",
        "\n",
        "\n",
        "# standardize\n",
        "vars = [\"gfs_tmin\", \"gfs_tmax\", \"us\"]\n",
        "for var in vars:\n",
        "  x_mean =x_train[var].mean()\n",
        "  x_std = np.std(x_train[var])\n",
        "\n",
        "  x_train[var] = (x_train[var] - x_mean) / x_std\n",
        "  x_val[var] = (x_val[var] - x_mean) / x_std\n",
        "  x_tst[var] = (x_tst[var] - x_mean) / x_std\n",
        "  x_all[var] = (x_all[var] - x_mean) / x_std\n",
        "\n",
        "\n",
        "y_vars = [\"tmin_K\", \"tmax_K\"]\n",
        "for var in y_vars:\n",
        "  y_mean =y_train[var].mean()\n",
        "  y_std = np.std(y_train[var])\n",
        "  y_train[var] = (y_train[var] - y_mean) / y_std\n",
        "  y_val[var] = (y_val[var] - y_mean) / y_std\n",
        "  y_tst[var] = (y_tst[var] - y_mean) / y_std\n",
        "  y_all[var] = (y_all[var] - y_mean) / y_std\n",
        "\n",
        "\n",
        "print(x_train.head(3))\n",
        "print(y_train.head(3))\n",
        "\n",
        "print(x_val.head(3))\n",
        "print(y_val.head(3))\n",
        "\n",
        "print(x_tst.head(3))\n",
        "print(y_tst.head(3))\n",
        "\n",
        "print(x_all.head(3))\n",
        "print(y_all.head(3))\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "x_tst.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lO_6KW63r1K3",
        "outputId": "9cebc3c7-e505-4380-a112-ad0520772693"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(432, 5)"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UIbG_uhOBQWx"
      },
      "source": [
        "## MLP with LAT/LON"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VZNkuiHRBaTK"
      },
      "source": [
        "### Hyperparams"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "nwYfmUGkBV0z"
      },
      "outputs": [],
      "source": [
        "# Define your hyperparameters and RMSE\n",
        "hyperparams = {\n",
        "    'solver': 'adam',\n",
        "    'activation': 'relu',\n",
        "    'alpha': 0.01, # L2-reg\n",
        "    'batch_size': 'auto',\n",
        "    'learning_rate_init': 0.01,\n",
        "    'learning_rate': 'adaptive',\n",
        "    'shuffling': True,\n",
        "    'validation_fraction': 0.2,\n",
        "    'num_models': 30,\n",
        "    'min_nhn': 10,\n",
        "    'max_nhn': 50,\n",
        "    'max_hidden_layers': 5,\n",
        "    'rmse': 0,\n",
        "    'ensemble_guid': ''\n",
        "}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_qpBMf2nBx5D",
        "outputId": "58a6cd39-f5a7-429c-c6cc-39893c788792"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[1;30;43mStreaming output truncated to the last 5000 lines.\u001b[0m\n",
            "tmax_K    0.632216\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.982530\n",
            "tmax_K    0.634109\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.971157\n",
            "tmax_K    0.633191\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.981067\n",
            "tmax_K    0.636719\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.976642\n",
            "tmax_K    0.632709\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.967927\n",
            "tmax_K    0.631474\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.980152\n",
            "tmax_K    0.633779\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.972131\n",
            "tmax_K    0.633715\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.986288\n",
            "tmax_K    0.636733\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.980900\n",
            "tmax_K    0.636836\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.966093\n",
            "tmax_K    0.629382\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.973214\n",
            "tmax_K    0.633809\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.975298\n",
            "tmax_K    0.633340\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.979670\n",
            "tmax_K    0.632316\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.979308\n",
            "tmax_K    0.637396\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.975215\n",
            "tmax_K    0.634203\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.990942\n",
            "tmax_K    0.638346\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.976722\n",
            "tmax_K    0.633387\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 2\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.987551\n",
            "tmax_K    0.632896\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.977664\n",
            "tmax_K    0.633182\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.970170\n",
            "tmax_K    0.631901\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.437854\n",
            "tmax_K    0.362189\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.980015\n",
            "tmax_K    0.632697\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.404170\n",
            "tmax_K    0.339193\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.985058\n",
            "tmax_K    0.636431\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.976816\n",
            "tmax_K    0.629344\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.429238\n",
            "tmax_K    0.366051\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.314758\n",
            "tmax_K    0.349814\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.977242\n",
            "tmax_K    0.634682\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.977109\n",
            "tmax_K    0.634659\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.968681\n",
            "tmax_K    0.630245\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.972492\n",
            "tmax_K    0.634571\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.977839\n",
            "tmax_K    0.633140\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.976247\n",
            "tmax_K    0.633519\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.987899\n",
            "tmax_K    0.633346\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.973747\n",
            "tmax_K    0.634048\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.989774\n",
            "tmax_K    0.637065\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.978306\n",
            "tmax_K    0.628777\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.978072\n",
            "tmax_K    0.631464\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.967674\n",
            "tmax_K    0.632249\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.976760\n",
            "tmax_K    0.631423\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.975369\n",
            "tmax_K    0.626645\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.353385\n",
            "tmax_K    0.363356\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.990350\n",
            "tmax_K    0.636454\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.988585\n",
            "tmax_K    0.635960\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.979014\n",
            "tmax_K    0.633937\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.980027\n",
            "tmax_K    0.634456\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.982992\n",
            "tmax_K    0.637570\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.979626\n",
            "tmax_K    0.632135\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.971857\n",
            "tmax_K    0.633351\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.986037\n",
            "tmax_K    0.636027\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.340980\n",
            "tmax_K    0.328454\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.965324\n",
            "tmax_K    0.626076\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.990734\n",
            "tmax_K    0.640637\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.975014\n",
            "tmax_K    0.633857\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.446456\n",
            "tmax_K    0.364913\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.359711\n",
            "tmax_K    0.339473\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.377207\n",
            "tmax_K    0.273016\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.978835\n",
            "tmax_K    0.634334\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 3\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.424290\n",
            "tmax_K    0.326587\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.969279\n",
            "tmax_K    0.631623\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.450136\n",
            "tmax_K    0.362779\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.428428\n",
            "tmax_K    0.383271\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.977244\n",
            "tmax_K    0.632114\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.478212\n",
            "tmax_K    0.352606\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.321010\n",
            "tmax_K    0.338468\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.979316\n",
            "tmax_K    0.634702\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.351504\n",
            "tmax_K    0.337117\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.975246\n",
            "tmax_K    0.632384\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.352071\n",
            "tmax_K    0.318726\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.979119\n",
            "tmax_K    0.631988\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.975768\n",
            "tmax_K    0.632173\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.301793\n",
            "tmax_K    0.262108\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.973985\n",
            "tmax_K    0.632124\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.973208\n",
            "tmax_K    0.630504\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.974199\n",
            "tmax_K    0.631338\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.978299\n",
            "tmax_K    0.634490\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.984978\n",
            "tmax_K    0.638096\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.338889\n",
            "tmax_K    0.346490\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.974934\n",
            "tmax_K    0.633260\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.978334\n",
            "tmax_K    0.633382\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.967435\n",
            "tmax_K    0.628471\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.982007\n",
            "tmax_K    0.636507\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.424631\n",
            "tmax_K    0.369612\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.975002\n",
            "tmax_K    0.634962\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.440064\n",
            "tmax_K    0.371172\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.976741\n",
            "tmax_K    0.634146\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.977798\n",
            "tmax_K    0.633852\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.978832\n",
            "tmax_K    0.633824\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.970831\n",
            "tmax_K    0.630433\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.978481\n",
            "tmax_K    0.634115\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.980320\n",
            "tmax_K    0.634246\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.431061\n",
            "tmax_K    0.370029\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.971632\n",
            "tmax_K    0.622362\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.980479\n",
            "tmax_K    0.638268\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.978741\n",
            "tmax_K    0.633528\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.355398\n",
            "tmax_K    0.330724\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.976001\n",
            "tmax_K    0.630771\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.969746\n",
            "tmax_K    0.630634\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.983972\n",
            "tmax_K    0.637152\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 4\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.440760\n",
            "tmax_K    0.372691\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.460301\n",
            "tmax_K    0.326295\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.983383\n",
            "tmax_K    0.632445\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.975644\n",
            "tmax_K    0.636897\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.444782\n",
            "tmax_K    0.359002\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.978784\n",
            "tmax_K    0.632619\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.978344\n",
            "tmax_K    0.633423\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.355045\n",
            "tmax_K    0.345524\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.437220\n",
            "tmax_K    0.361175\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.398935\n",
            "tmax_K    0.369204\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.987185\n",
            "tmax_K    0.632798\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.324851\n",
            "tmax_K    0.268020\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.325483\n",
            "tmax_K    0.299663\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.968273\n",
            "tmax_K    0.633788\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.477408\n",
            "tmax_K    0.346501\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.377938\n",
            "tmax_K    0.308449\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.976524\n",
            "tmax_K    0.632191\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.977878\n",
            "tmax_K    0.633977\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.972571\n",
            "tmax_K    0.635252\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.406731\n",
            "tmax_K    0.347615\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.398788\n",
            "tmax_K    0.332382\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.309813\n",
            "tmax_K    0.321208\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.976713\n",
            "tmax_K    0.629387\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.443684\n",
            "tmax_K    0.326736\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.978981\n",
            "tmax_K    0.633999\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.977725\n",
            "tmax_K    0.632644\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.980132\n",
            "tmax_K    0.635463\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.973967\n",
            "tmax_K    0.633178\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.976943\n",
            "tmax_K    0.630936\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.974686\n",
            "tmax_K    0.631776\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.449308\n",
            "tmax_K    0.369359\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.976970\n",
            "tmax_K    0.639069\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.343898\n",
            "tmax_K    0.313432\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.976147\n",
            "tmax_K    0.630419\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.450523\n",
            "tmax_K    0.332196\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.318771\n",
            "tmax_K    0.343990\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.978146\n",
            "tmax_K    0.634023\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.328826\n",
            "tmax_K    0.331024\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.978722\n",
            "tmax_K    0.630557\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.981824\n",
            "tmax_K    0.640716\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.966989\n",
            "tmax_K    0.628917\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 5\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.465246\n",
            "tmax_K    0.377327\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.463181\n",
            "tmax_K    0.317366\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.311983\n",
            "tmax_K    0.328710\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.331742\n",
            "tmax_K    0.385360\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.360818\n",
            "tmax_K    0.258788\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.351771\n",
            "tmax_K    0.329979\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.257984\n",
            "tmax_K    0.291628\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.317105\n",
            "tmax_K    0.302871\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.994296\n",
            "tmax_K    0.634438\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.313014\n",
            "tmax_K    0.270083\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.341486\n",
            "tmax_K    0.340300\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.339979\n",
            "tmax_K    0.348018\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.975968\n",
            "tmax_K    0.634875\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.495670\n",
            "tmax_K    0.337843\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.291969\n",
            "tmax_K    0.328585\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.379538\n",
            "tmax_K    0.342794\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.975930\n",
            "tmax_K    0.632822\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.977548\n",
            "tmax_K    0.632647\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.333983\n",
            "tmax_K    0.287913\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.980821\n",
            "tmax_K    0.636748\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.981735\n",
            "tmax_K    0.633760\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.319356\n",
            "tmax_K    0.274089\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.302787\n",
            "tmax_K    0.354007\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.444691\n",
            "tmax_K    0.361391\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.978611\n",
            "tmax_K    0.634392\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.278425\n",
            "tmax_K    0.371117\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.376304\n",
            "tmax_K    0.317031\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.319664\n",
            "tmax_K    0.318723\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.976408\n",
            "tmax_K    0.632488\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.371871\n",
            "tmax_K    0.308015\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.973008\n",
            "tmax_K    0.636199\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.448671\n",
            "tmax_K    0.355629\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.982663\n",
            "tmax_K    0.634948\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.352826\n",
            "tmax_K    0.348229\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.359592\n",
            "tmax_K    0.322603\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.462837\n",
            "tmax_K    0.349421\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.331695\n",
            "tmax_K    0.305446\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.979605\n",
            "tmax_K    0.634051\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.979199\n",
            "tmax_K    0.633884\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.975490\n",
            "tmax_K    0.632789\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.970610\n",
            "tmax_K    0.635072\n",
            "dtype: float64\n",
            "[1.609488   1.62460028 1.60590943 0.80887202 0.71485071 1.6093419\n",
            " 1.62243893 1.59824222 1.61175198 1.61598717 1.6131082  1.61368421\n",
            " 0.71462242 1.60715334 1.63268797 1.60151235 1.6066082  1.62650785\n",
            " 1.59387828 0.70776693 1.61745573 1.60817354 1.62046169 1.61401845\n",
            " 1.61663925 1.60434768 1.61778645 1.60935056 1.59940066 1.61393081\n",
            " 1.60584603 1.62302095 1.61773645 1.59547438 1.60702362 1.60863835\n",
            " 1.61198565 1.61670387 1.60941814 1.62928807 1.61010858 1.62044789\n",
            " 1.61084565 1.60207088 0.80004339 1.61271263 0.74336293 1.62148902\n",
            " 1.6061594  0.79528856 0.66457177 1.61192337 1.61176829 1.59892534\n",
            " 1.60706304 1.61097947 1.60976653 1.62124438 1.60779564 1.62683931\n",
            " 1.60708335 1.60953632 1.59992234 1.60818327 1.60201398 0.71674073\n",
            " 1.6268035  1.62454557 1.61295106 1.61448239 1.62056216 1.61176084\n",
            " 1.60520829 1.62206354 0.66943386 1.5913998  1.63137142 1.60887051\n",
            " 0.81136882 0.69918398 0.65022239 1.61316906 0.75087645 1.60090249\n",
            " 0.81291476 0.81169944 1.6093584  0.83081718 0.65947736 1.61401831\n",
            " 0.6886201  1.60763038 0.67079753 1.61110712 1.60794051 0.56390102\n",
            " 1.60610876 1.60371145 1.60553716 1.61278873 1.62307338 0.6853792\n",
            " 1.60819342 1.61171616 1.59590596 1.61851382 0.79424327 1.60996389\n",
            " 0.81123624 1.61088694 1.61164987 1.61265581 1.60126457 1.61259582\n",
            " 1.61456592 0.80108951 1.59399414 1.61874661 1.61226939 0.68612125\n",
            " 1.60677151 1.60037987 1.62112333 0.81345107 0.78659632 1.61582845\n",
            " 1.6125403  0.8037839  1.61140245 1.61176684 0.70056946 0.79839503\n",
            " 0.76813833 1.6199831  0.59287121 0.62514576 1.60206031 0.82390891\n",
            " 0.68638694 1.60871496 1.61185534 1.60782303 0.75434633 0.73116978\n",
            " 0.63102095 1.60609989 0.77042046 1.61298064 1.61036947 1.61559483\n",
            " 1.60714561 1.60787965 1.6064616  0.81866691 1.61603886 0.65732991\n",
            " 1.60656675 0.78271903 0.66276136 1.61216927 0.65984995 1.60927953\n",
            " 1.62253946 1.59590585 0.84257245 0.78054659 0.64069349 0.71710207\n",
            " 0.61960666 0.68175009 0.54961218 0.61997673 1.6287343  0.58309677\n",
            " 0.68178549 0.68799701 1.61084362 0.83351228 0.62055415 0.72233234\n",
            " 1.60875178 1.61019508 0.62189563 1.61756874 1.61549501 0.59344472\n",
            " 0.65679469 0.8060813  1.61300333 0.64954248 0.69333539 0.6383877\n",
            " 1.60889541 0.67988651 1.60920758 0.80429927 1.61761167 0.70105514\n",
            " 0.68219475 0.81225796 0.63714048 1.61365631 1.61308302 1.60827856\n",
            " 1.60568237]\n",
            "[tmin_K    0.977905\n",
            "tmax_K    0.631583\n",
            "dtype: float64, tmin_K    0.989254\n",
            "tmax_K    0.635346\n",
            "dtype: float64, tmin_K    0.970357\n",
            "tmax_K    0.635552\n",
            "dtype: float64, tmin_K    0.448491\n",
            "tmax_K    0.360382\n",
            "dtype: float64, tmin_K    0.377268\n",
            "tmax_K    0.337583\n",
            "dtype: float64, tmin_K    0.976968\n",
            "tmax_K    0.632374\n",
            "dtype: float64, tmin_K    0.981604\n",
            "tmax_K    0.640835\n",
            "dtype: float64, tmin_K    0.970247\n",
            "tmax_K    0.627995\n",
            "dtype: float64, tmin_K    0.977512\n",
            "tmax_K    0.634240\n",
            "dtype: float64, tmin_K    0.980739\n",
            "tmax_K    0.635248\n",
            "dtype: float64, tmin_K    0.979860\n",
            "tmax_K    0.633249\n",
            "dtype: float64, tmin_K    0.980207\n",
            "tmax_K    0.633477\n",
            "dtype: float64, tmin_K    0.352421\n",
            "tmax_K    0.362201\n",
            "dtype: float64, tmin_K    0.974478\n",
            "tmax_K    0.632675\n",
            "dtype: float64, tmin_K    0.992571\n",
            "tmax_K    0.640117\n",
            "dtype: float64, tmin_K    0.968740\n",
            "tmax_K    0.632772\n",
            "dtype: float64, tmin_K    0.971906\n",
            "tmax_K    0.634702\n",
            "dtype: float64, tmin_K    0.989475\n",
            "tmax_K    0.637032\n",
            "dtype: float64, tmin_K    0.960662\n",
            "tmax_K    0.633216\n",
            "dtype: float64, tmin_K    0.364492\n",
            "tmax_K    0.343274\n",
            "dtype: float64, tmin_K    0.980203\n",
            "tmax_K    0.637253\n",
            "dtype: float64, tmin_K    0.974270\n",
            "tmax_K    0.633903\n",
            "dtype: float64, tmin_K    0.983534\n",
            "tmax_K    0.636928\n",
            "dtype: float64, tmin_K    0.981803\n",
            "tmax_K    0.632216\n",
            "dtype: float64, tmin_K    0.982530\n",
            "tmax_K    0.634109\n",
            "dtype: float64, tmin_K    0.971157\n",
            "tmax_K    0.633191\n",
            "dtype: float64, tmin_K    0.981067\n",
            "tmax_K    0.636719\n",
            "dtype: float64, tmin_K    0.976642\n",
            "tmax_K    0.632709\n",
            "dtype: float64, tmin_K    0.967927\n",
            "tmax_K    0.631474\n",
            "dtype: float64, tmin_K    0.980152\n",
            "tmax_K    0.633779\n",
            "dtype: float64, tmin_K    0.972131\n",
            "tmax_K    0.633715\n",
            "dtype: float64, tmin_K    0.986288\n",
            "tmax_K    0.636733\n",
            "dtype: float64, tmin_K    0.980900\n",
            "tmax_K    0.636836\n",
            "dtype: float64, tmin_K    0.966093\n",
            "tmax_K    0.629382\n",
            "dtype: float64, tmin_K    0.973214\n",
            "tmax_K    0.633809\n",
            "dtype: float64, tmin_K    0.975298\n",
            "tmax_K    0.633340\n",
            "dtype: float64, tmin_K    0.979670\n",
            "tmax_K    0.632316\n",
            "dtype: float64, tmin_K    0.979308\n",
            "tmax_K    0.637396\n",
            "dtype: float64, tmin_K    0.975215\n",
            "tmax_K    0.634203\n",
            "dtype: float64, tmin_K    0.990942\n",
            "tmax_K    0.638346\n",
            "dtype: float64, tmin_K    0.976722\n",
            "tmax_K    0.633387\n",
            "dtype: float64, tmin_K    0.987551\n",
            "tmax_K    0.632896\n",
            "dtype: float64, tmin_K    0.977664\n",
            "tmax_K    0.633182\n",
            "dtype: float64, tmin_K    0.970170\n",
            "tmax_K    0.631901\n",
            "dtype: float64, tmin_K    0.437854\n",
            "tmax_K    0.362189\n",
            "dtype: float64, tmin_K    0.980015\n",
            "tmax_K    0.632697\n",
            "dtype: float64, tmin_K    0.404170\n",
            "tmax_K    0.339193\n",
            "dtype: float64, tmin_K    0.985058\n",
            "tmax_K    0.636431\n",
            "dtype: float64, tmin_K    0.976816\n",
            "tmax_K    0.629344\n",
            "dtype: float64, tmin_K    0.429238\n",
            "tmax_K    0.366051\n",
            "dtype: float64, tmin_K    0.314758\n",
            "tmax_K    0.349814\n",
            "dtype: float64, tmin_K    0.977242\n",
            "tmax_K    0.634682\n",
            "dtype: float64, tmin_K    0.977109\n",
            "tmax_K    0.634659\n",
            "dtype: float64, tmin_K    0.968681\n",
            "tmax_K    0.630245\n",
            "dtype: float64, tmin_K    0.972492\n",
            "tmax_K    0.634571\n",
            "dtype: float64, tmin_K    0.977839\n",
            "tmax_K    0.633140\n",
            "dtype: float64, tmin_K    0.976247\n",
            "tmax_K    0.633519\n",
            "dtype: float64, tmin_K    0.987899\n",
            "tmax_K    0.633346\n",
            "dtype: float64, tmin_K    0.973747\n",
            "tmax_K    0.634048\n",
            "dtype: float64, tmin_K    0.989774\n",
            "tmax_K    0.637065\n",
            "dtype: float64, tmin_K    0.978306\n",
            "tmax_K    0.628777\n",
            "dtype: float64, tmin_K    0.978072\n",
            "tmax_K    0.631464\n",
            "dtype: float64, tmin_K    0.967674\n",
            "tmax_K    0.632249\n",
            "dtype: float64, tmin_K    0.976760\n",
            "tmax_K    0.631423\n",
            "dtype: float64, tmin_K    0.975369\n",
            "tmax_K    0.626645\n",
            "dtype: float64, tmin_K    0.353385\n",
            "tmax_K    0.363356\n",
            "dtype: float64, tmin_K    0.990350\n",
            "tmax_K    0.636454\n",
            "dtype: float64, tmin_K    0.988585\n",
            "tmax_K    0.635960\n",
            "dtype: float64, tmin_K    0.979014\n",
            "tmax_K    0.633937\n",
            "dtype: float64, tmin_K    0.980027\n",
            "tmax_K    0.634456\n",
            "dtype: float64, tmin_K    0.982992\n",
            "tmax_K    0.637570\n",
            "dtype: float64, tmin_K    0.979626\n",
            "tmax_K    0.632135\n",
            "dtype: float64, tmin_K    0.971857\n",
            "tmax_K    0.633351\n",
            "dtype: float64, tmin_K    0.986037\n",
            "tmax_K    0.636027\n",
            "dtype: float64, tmin_K    0.340980\n",
            "tmax_K    0.328454\n",
            "dtype: float64, tmin_K    0.965324\n",
            "tmax_K    0.626076\n",
            "dtype: float64, tmin_K    0.990734\n",
            "tmax_K    0.640637\n",
            "dtype: float64, tmin_K    0.975014\n",
            "tmax_K    0.633857\n",
            "dtype: float64, tmin_K    0.446456\n",
            "tmax_K    0.364913\n",
            "dtype: float64, tmin_K    0.359711\n",
            "tmax_K    0.339473\n",
            "dtype: float64, tmin_K    0.377207\n",
            "tmax_K    0.273016\n",
            "dtype: float64, tmin_K    0.978835\n",
            "tmax_K    0.634334\n",
            "dtype: float64, tmin_K    0.424290\n",
            "tmax_K    0.326587\n",
            "dtype: float64, tmin_K    0.969279\n",
            "tmax_K    0.631623\n",
            "dtype: float64, tmin_K    0.450136\n",
            "tmax_K    0.362779\n",
            "dtype: float64, tmin_K    0.428428\n",
            "tmax_K    0.383271\n",
            "dtype: float64, tmin_K    0.977244\n",
            "tmax_K    0.632114\n",
            "dtype: float64, tmin_K    0.478212\n",
            "tmax_K    0.352606\n",
            "dtype: float64, tmin_K    0.321010\n",
            "tmax_K    0.338468\n",
            "dtype: float64, tmin_K    0.979316\n",
            "tmax_K    0.634702\n",
            "dtype: float64, tmin_K    0.351504\n",
            "tmax_K    0.337117\n",
            "dtype: float64, tmin_K    0.975246\n",
            "tmax_K    0.632384\n",
            "dtype: float64, tmin_K    0.352071\n",
            "tmax_K    0.318726\n",
            "dtype: float64, tmin_K    0.979119\n",
            "tmax_K    0.631988\n",
            "dtype: float64, tmin_K    0.975768\n",
            "tmax_K    0.632173\n",
            "dtype: float64, tmin_K    0.301793\n",
            "tmax_K    0.262108\n",
            "dtype: float64, tmin_K    0.973985\n",
            "tmax_K    0.632124\n",
            "dtype: float64, tmin_K    0.973208\n",
            "tmax_K    0.630504\n",
            "dtype: float64, tmin_K    0.974199\n",
            "tmax_K    0.631338\n",
            "dtype: float64, tmin_K    0.978299\n",
            "tmax_K    0.634490\n",
            "dtype: float64, tmin_K    0.984978\n",
            "tmax_K    0.638096\n",
            "dtype: float64, tmin_K    0.338889\n",
            "tmax_K    0.346490\n",
            "dtype: float64, tmin_K    0.974934\n",
            "tmax_K    0.633260\n",
            "dtype: float64, tmin_K    0.978334\n",
            "tmax_K    0.633382\n",
            "dtype: float64, tmin_K    0.967435\n",
            "tmax_K    0.628471\n",
            "dtype: float64, tmin_K    0.982007\n",
            "tmax_K    0.636507\n",
            "dtype: float64, tmin_K    0.424631\n",
            "tmax_K    0.369612\n",
            "dtype: float64, tmin_K    0.975002\n",
            "tmax_K    0.634962\n",
            "dtype: float64, tmin_K    0.440064\n",
            "tmax_K    0.371172\n",
            "dtype: float64, tmin_K    0.976741\n",
            "tmax_K    0.634146\n",
            "dtype: float64, tmin_K    0.977798\n",
            "tmax_K    0.633852\n",
            "dtype: float64, tmin_K    0.978832\n",
            "tmax_K    0.633824\n",
            "dtype: float64, tmin_K    0.970831\n",
            "tmax_K    0.630433\n",
            "dtype: float64, tmin_K    0.978481\n",
            "tmax_K    0.634115\n",
            "dtype: float64, tmin_K    0.980320\n",
            "tmax_K    0.634246\n",
            "dtype: float64, tmin_K    0.431061\n",
            "tmax_K    0.370029\n",
            "dtype: float64, tmin_K    0.971632\n",
            "tmax_K    0.622362\n",
            "dtype: float64, tmin_K    0.980479\n",
            "tmax_K    0.638268\n",
            "dtype: float64, tmin_K    0.978741\n",
            "tmax_K    0.633528\n",
            "dtype: float64, tmin_K    0.355398\n",
            "tmax_K    0.330724\n",
            "dtype: float64, tmin_K    0.976001\n",
            "tmax_K    0.630771\n",
            "dtype: float64, tmin_K    0.969746\n",
            "tmax_K    0.630634\n",
            "dtype: float64, tmin_K    0.983972\n",
            "tmax_K    0.637152\n",
            "dtype: float64, tmin_K    0.440760\n",
            "tmax_K    0.372691\n",
            "dtype: float64, tmin_K    0.460301\n",
            "tmax_K    0.326295\n",
            "dtype: float64, tmin_K    0.983383\n",
            "tmax_K    0.632445\n",
            "dtype: float64, tmin_K    0.975644\n",
            "tmax_K    0.636897\n",
            "dtype: float64, tmin_K    0.444782\n",
            "tmax_K    0.359002\n",
            "dtype: float64, tmin_K    0.978784\n",
            "tmax_K    0.632619\n",
            "dtype: float64, tmin_K    0.978344\n",
            "tmax_K    0.633423\n",
            "dtype: float64, tmin_K    0.355045\n",
            "tmax_K    0.345524\n",
            "dtype: float64, tmin_K    0.437220\n",
            "tmax_K    0.361175\n",
            "dtype: float64, tmin_K    0.398935\n",
            "tmax_K    0.369204\n",
            "dtype: float64, tmin_K    0.987185\n",
            "tmax_K    0.632798\n",
            "dtype: float64, tmin_K    0.324851\n",
            "tmax_K    0.268020\n",
            "dtype: float64, tmin_K    0.325483\n",
            "tmax_K    0.299663\n",
            "dtype: float64, tmin_K    0.968273\n",
            "tmax_K    0.633788\n",
            "dtype: float64, tmin_K    0.477408\n",
            "tmax_K    0.346501\n",
            "dtype: float64, tmin_K    0.377938\n",
            "tmax_K    0.308449\n",
            "dtype: float64, tmin_K    0.976524\n",
            "tmax_K    0.632191\n",
            "dtype: float64, tmin_K    0.977878\n",
            "tmax_K    0.633977\n",
            "dtype: float64, tmin_K    0.972571\n",
            "tmax_K    0.635252\n",
            "dtype: float64, tmin_K    0.406731\n",
            "tmax_K    0.347615\n",
            "dtype: float64, tmin_K    0.398788\n",
            "tmax_K    0.332382\n",
            "dtype: float64, tmin_K    0.309813\n",
            "tmax_K    0.321208\n",
            "dtype: float64, tmin_K    0.976713\n",
            "tmax_K    0.629387\n",
            "dtype: float64, tmin_K    0.443684\n",
            "tmax_K    0.326736\n",
            "dtype: float64, tmin_K    0.978981\n",
            "tmax_K    0.633999\n",
            "dtype: float64, tmin_K    0.977725\n",
            "tmax_K    0.632644\n",
            "dtype: float64, tmin_K    0.980132\n",
            "tmax_K    0.635463\n",
            "dtype: float64, tmin_K    0.973967\n",
            "tmax_K    0.633178\n",
            "dtype: float64, tmin_K    0.976943\n",
            "tmax_K    0.630936\n",
            "dtype: float64, tmin_K    0.974686\n",
            "tmax_K    0.631776\n",
            "dtype: float64, tmin_K    0.449308\n",
            "tmax_K    0.369359\n",
            "dtype: float64, tmin_K    0.976970\n",
            "tmax_K    0.639069\n",
            "dtype: float64, tmin_K    0.343898\n",
            "tmax_K    0.313432\n",
            "dtype: float64, tmin_K    0.976147\n",
            "tmax_K    0.630419\n",
            "dtype: float64, tmin_K    0.450523\n",
            "tmax_K    0.332196\n",
            "dtype: float64, tmin_K    0.318771\n",
            "tmax_K    0.343990\n",
            "dtype: float64, tmin_K    0.978146\n",
            "tmax_K    0.634023\n",
            "dtype: float64, tmin_K    0.328826\n",
            "tmax_K    0.331024\n",
            "dtype: float64, tmin_K    0.978722\n",
            "tmax_K    0.630557\n",
            "dtype: float64, tmin_K    0.981824\n",
            "tmax_K    0.640716\n",
            "dtype: float64, tmin_K    0.966989\n",
            "tmax_K    0.628917\n",
            "dtype: float64, tmin_K    0.465246\n",
            "tmax_K    0.377327\n",
            "dtype: float64, tmin_K    0.463181\n",
            "tmax_K    0.317366\n",
            "dtype: float64, tmin_K    0.311983\n",
            "tmax_K    0.328710\n",
            "dtype: float64, tmin_K    0.331742\n",
            "tmax_K    0.385360\n",
            "dtype: float64, tmin_K    0.360818\n",
            "tmax_K    0.258788\n",
            "dtype: float64, tmin_K    0.351771\n",
            "tmax_K    0.329979\n",
            "dtype: float64, tmin_K    0.257984\n",
            "tmax_K    0.291628\n",
            "dtype: float64, tmin_K    0.317105\n",
            "tmax_K    0.302871\n",
            "dtype: float64, tmin_K    0.994296\n",
            "tmax_K    0.634438\n",
            "dtype: float64, tmin_K    0.313014\n",
            "tmax_K    0.270083\n",
            "dtype: float64, tmin_K    0.341486\n",
            "tmax_K    0.340300\n",
            "dtype: float64, tmin_K    0.339979\n",
            "tmax_K    0.348018\n",
            "dtype: float64, tmin_K    0.975968\n",
            "tmax_K    0.634875\n",
            "dtype: float64, tmin_K    0.495670\n",
            "tmax_K    0.337843\n",
            "dtype: float64, tmin_K    0.291969\n",
            "tmax_K    0.328585\n",
            "dtype: float64, tmin_K    0.379538\n",
            "tmax_K    0.342794\n",
            "dtype: float64, tmin_K    0.975930\n",
            "tmax_K    0.632822\n",
            "dtype: float64, tmin_K    0.977548\n",
            "tmax_K    0.632647\n",
            "dtype: float64, tmin_K    0.333983\n",
            "tmax_K    0.287913\n",
            "dtype: float64, tmin_K    0.980821\n",
            "tmax_K    0.636748\n",
            "dtype: float64, tmin_K    0.981735\n",
            "tmax_K    0.633760\n",
            "dtype: float64, tmin_K    0.319356\n",
            "tmax_K    0.274089\n",
            "dtype: float64, tmin_K    0.302787\n",
            "tmax_K    0.354007\n",
            "dtype: float64, tmin_K    0.444691\n",
            "tmax_K    0.361391\n",
            "dtype: float64, tmin_K    0.978611\n",
            "tmax_K    0.634392\n",
            "dtype: float64, tmin_K    0.278425\n",
            "tmax_K    0.371117\n",
            "dtype: float64, tmin_K    0.376304\n",
            "tmax_K    0.317031\n",
            "dtype: float64, tmin_K    0.319664\n",
            "tmax_K    0.318723\n",
            "dtype: float64, tmin_K    0.976408\n",
            "tmax_K    0.632488\n",
            "dtype: float64, tmin_K    0.371871\n",
            "tmax_K    0.308015\n",
            "dtype: float64, tmin_K    0.973008\n",
            "tmax_K    0.636199\n",
            "dtype: float64, tmin_K    0.448671\n",
            "tmax_K    0.355629\n",
            "dtype: float64, tmin_K    0.982663\n",
            "tmax_K    0.634948\n",
            "dtype: float64, tmin_K    0.352826\n",
            "tmax_K    0.348229\n",
            "dtype: float64, tmin_K    0.359592\n",
            "tmax_K    0.322603\n",
            "dtype: float64, tmin_K    0.462837\n",
            "tmax_K    0.349421\n",
            "dtype: float64, tmin_K    0.331695\n",
            "tmax_K    0.305446\n",
            "dtype: float64, tmin_K    0.979605\n",
            "tmax_K    0.634051\n",
            "dtype: float64, tmin_K    0.979199\n",
            "tmax_K    0.633884\n",
            "dtype: float64, tmin_K    0.975490\n",
            "tmax_K    0.632789\n",
            "dtype: float64, tmin_K    0.970610\n",
            "tmax_K    0.635072\n",
            "dtype: float64]\n",
            "\t BEST: 5 hidden layers, 16 hidden neurons\n",
            "Model Number: 27\n",
            "\t # Hidden Layers = 1\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.985936\n",
            "tmax_K    0.635460\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.978321\n",
            "tmax_K    0.633645\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.972273\n",
            "tmax_K    0.632902\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.982850\n",
            "tmax_K    0.637208\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.980818\n",
            "tmax_K    0.636320\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.977420\n",
            "tmax_K    0.633492\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.972225\n",
            "tmax_K    0.632465\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.977667\n",
            "tmax_K    0.634004\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.977879\n",
            "tmax_K    0.631982\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.980284\n",
            "tmax_K    0.634369\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.980861\n",
            "tmax_K    0.636858\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.964608\n",
            "tmax_K    0.630473\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.975590\n",
            "tmax_K    0.630059\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.983733\n",
            "tmax_K    0.637679\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.978138\n",
            "tmax_K    0.631790\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.985373\n",
            "tmax_K    0.636630\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.984500\n",
            "tmax_K    0.635013\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.983826\n",
            "tmax_K    0.638332\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.977124\n",
            "tmax_K    0.630380\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.978319\n",
            "tmax_K    0.633265\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.969590\n",
            "tmax_K    0.627903\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.986040\n",
            "tmax_K    0.631556\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.980267\n",
            "tmax_K    0.633703\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.994150\n",
            "tmax_K    0.636666\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.974650\n",
            "tmax_K    0.632589\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.974478\n",
            "tmax_K    0.632552\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.976484\n",
            "tmax_K    0.634202\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.988475\n",
            "tmax_K    0.638013\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.975566\n",
            "tmax_K    0.632143\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.978796\n",
            "tmax_K    0.633823\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.350832\n",
            "tmax_K    0.349963\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.975594\n",
            "tmax_K    0.631557\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.977960\n",
            "tmax_K    0.633736\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.968848\n",
            "tmax_K    0.634118\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.985032\n",
            "tmax_K    0.635177\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.977362\n",
            "tmax_K    0.631087\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.972711\n",
            "tmax_K    0.626214\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.977995\n",
            "tmax_K    0.635404\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.975061\n",
            "tmax_K    0.631656\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.988590\n",
            "tmax_K    0.641585\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.980289\n",
            "tmax_K    0.635325\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 2\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.977494\n",
            "tmax_K    0.634248\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.319726\n",
            "tmax_K    0.304479\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.972603\n",
            "tmax_K    0.634428\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.468067\n",
            "tmax_K    0.344525\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.338264\n",
            "tmax_K    0.339934\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.977127\n",
            "tmax_K    0.632039\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.975184\n",
            "tmax_K    0.638086\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.977268\n",
            "tmax_K    0.633545\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.980263\n",
            "tmax_K    0.634552\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.982250\n",
            "tmax_K    0.636144\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.987948\n",
            "tmax_K    0.633313\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.338732\n",
            "tmax_K    0.309333\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.976781\n",
            "tmax_K    0.633443\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.980163\n",
            "tmax_K    0.634608\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.345709\n",
            "tmax_K    0.360556\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.974869\n",
            "tmax_K    0.632828\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.977450\n",
            "tmax_K    0.635153\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.970475\n",
            "tmax_K    0.632432\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.978313\n",
            "tmax_K    0.635199\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.342550\n",
            "tmax_K    0.364233\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.978882\n",
            "tmax_K    0.634045\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.311480\n",
            "tmax_K    0.280635\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.983749\n",
            "tmax_K    0.637913\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.973017\n",
            "tmax_K    0.628579\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.987021\n",
            "tmax_K    0.634920\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.329672\n",
            "tmax_K    0.302710\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.460957\n",
            "tmax_K    0.360281\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.977793\n",
            "tmax_K    0.634160\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.980597\n",
            "tmax_K    0.633506\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.978344\n",
            "tmax_K    0.635211\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.982347\n",
            "tmax_K    0.634701\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.301418\n",
            "tmax_K    0.292964\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.974115\n",
            "tmax_K    0.631724\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.975666\n",
            "tmax_K    0.633997\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.980036\n",
            "tmax_K    0.633374\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.408986\n",
            "tmax_K    0.295725\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.967869\n",
            "tmax_K    0.634463\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.987167\n",
            "tmax_K    0.635954\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.977567\n",
            "tmax_K    0.632364\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.976943\n",
            "tmax_K    0.631058\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.990245\n",
            "tmax_K    0.639689\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 3\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.971165\n",
            "tmax_K    0.632985\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.979039\n",
            "tmax_K    0.629314\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.975337\n",
            "tmax_K    0.633049\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.967386\n",
            "tmax_K    0.631203\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.972237\n",
            "tmax_K    0.629924\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.411403\n",
            "tmax_K    0.361944\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.359713\n",
            "tmax_K    0.347515\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.977339\n",
            "tmax_K    0.633811\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.976947\n",
            "tmax_K    0.638870\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.372536\n",
            "tmax_K    0.319162\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.966826\n",
            "tmax_K    0.634347\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.436729\n",
            "tmax_K    0.352454\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.977487\n",
            "tmax_K    0.634166\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.981864\n",
            "tmax_K    0.633831\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.978381\n",
            "tmax_K    0.633985\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.979036\n",
            "tmax_K    0.634328\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.346998\n",
            "tmax_K    0.316833\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.970974\n",
            "tmax_K    0.632522\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.977956\n",
            "tmax_K    0.634272\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.970138\n",
            "tmax_K    0.631575\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.988762\n",
            "tmax_K    0.631044\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.990534\n",
            "tmax_K    0.633991\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.326670\n",
            "tmax_K    0.275278\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.978062\n",
            "tmax_K    0.633508\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.435321\n",
            "tmax_K    0.362302\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.984536\n",
            "tmax_K    0.632120\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.976186\n",
            "tmax_K    0.635767\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.980900\n",
            "tmax_K    0.634971\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.977507\n",
            "tmax_K    0.633257\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.970511\n",
            "tmax_K    0.632126\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.986632\n",
            "tmax_K    0.637903\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.350028\n",
            "tmax_K    0.316545\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.977561\n",
            "tmax_K    0.637130\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.977237\n",
            "tmax_K    0.633073\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.978519\n",
            "tmax_K    0.634384\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.976209\n",
            "tmax_K    0.632887\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.978424\n",
            "tmax_K    0.632735\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.982859\n",
            "tmax_K    0.638037\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.976910\n",
            "tmax_K    0.634352\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.978688\n",
            "tmax_K    0.633601\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.971325\n",
            "tmax_K    0.630981\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 4\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.977178\n",
            "tmax_K    0.633200\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.463280\n",
            "tmax_K    0.363739\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.436675\n",
            "tmax_K    0.345956\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.380981\n",
            "tmax_K    0.312902\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.449606\n",
            "tmax_K    0.359048\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.423237\n",
            "tmax_K    0.325153\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.983799\n",
            "tmax_K    0.639836\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.449419\n",
            "tmax_K    0.343337\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.982297\n",
            "tmax_K    0.634803\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.277394\n",
            "tmax_K    0.244181\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.392451\n",
            "tmax_K    0.312640\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.435825\n",
            "tmax_K    0.298333\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.978210\n",
            "tmax_K    0.634055\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.424146\n",
            "tmax_K    0.325113\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.337340\n",
            "tmax_K    0.355751\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.363606\n",
            "tmax_K    0.341713\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.358494\n",
            "tmax_K    0.284996\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.467204\n",
            "tmax_K    0.344671\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.975636\n",
            "tmax_K    0.632459\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.985993\n",
            "tmax_K    0.630265\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.980392\n",
            "tmax_K    0.634889\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.982248\n",
            "tmax_K    0.638048\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.454718\n",
            "tmax_K    0.347975\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.323566\n",
            "tmax_K    0.366910\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.976764\n",
            "tmax_K    0.633691\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.977178\n",
            "tmax_K    0.633296\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.977973\n",
            "tmax_K    0.634320\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.393268\n",
            "tmax_K    0.267299\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.344033\n",
            "tmax_K    0.271959\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.976173\n",
            "tmax_K    0.626027\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.345469\n",
            "tmax_K    0.366202\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.981843\n",
            "tmax_K    0.633871\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.335524\n",
            "tmax_K    0.342493\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.291814\n",
            "tmax_K    0.346421\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.982469\n",
            "tmax_K    0.635525\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.973704\n",
            "tmax_K    0.634794\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.977728\n",
            "tmax_K    0.629846\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.335358\n",
            "tmax_K    0.361059\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.976734\n",
            "tmax_K    0.634229\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.446506\n",
            "tmax_K    0.360210\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.373723\n",
            "tmax_K    0.323692\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 5\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    1.015269\n",
            "tmax_K    0.610938\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.460593\n",
            "tmax_K    0.380437\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.469710\n",
            "tmax_K    0.351133\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.370489\n",
            "tmax_K    0.318794\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.445201\n",
            "tmax_K    0.350667\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.342178\n",
            "tmax_K    0.371723\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.319326\n",
            "tmax_K    0.291460\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.268922\n",
            "tmax_K    0.251573\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.343855\n",
            "tmax_K    0.331658\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.368756\n",
            "tmax_K    0.230186\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.346131\n",
            "tmax_K    0.325509\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.389462\n",
            "tmax_K    0.280255\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.354533\n",
            "tmax_K    0.313638\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.977042\n",
            "tmax_K    0.630949\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.364255\n",
            "tmax_K    0.324294\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.342781\n",
            "tmax_K    0.315720\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.339056\n",
            "tmax_K    0.343104\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.378877\n",
            "tmax_K    0.349542\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.988280\n",
            "tmax_K    0.639127\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.985424\n",
            "tmax_K    0.636111\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.973728\n",
            "tmax_K    0.634477\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.313441\n",
            "tmax_K    0.260130\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.979336\n",
            "tmax_K    0.632461\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.435215\n",
            "tmax_K    0.350955\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.363963\n",
            "tmax_K    0.357784\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.972099\n",
            "tmax_K    0.634295\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.43363\n",
            "tmax_K    0.35728\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.368634\n",
            "tmax_K    0.299841\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.980416\n",
            "tmax_K    0.633077\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.982997\n",
            "tmax_K    0.634182\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.974954\n",
            "tmax_K    0.632716\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.977270\n",
            "tmax_K    0.635004\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.979443\n",
            "tmax_K    0.631260\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.975783\n",
            "tmax_K    0.640516\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.375129\n",
            "tmax_K    0.349366\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.979046\n",
            "tmax_K    0.634303\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.363352\n",
            "tmax_K    0.343933\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.978003\n",
            "tmax_K    0.637168\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.395366\n",
            "tmax_K    0.308170\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.978732\n",
            "tmax_K    0.635714\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.978648\n",
            "tmax_K    0.636209\n",
            "dtype: float64\n",
            "[1.62139551 1.61196581 1.60517473 1.62005781 1.61713829 1.61091225\n",
            " 1.60468993 1.61167138 1.60986046 1.61465314 1.61771858 1.59508119\n",
            " 1.60564941 1.62141112 1.60992823 1.62200357 1.61951364 1.62215836\n",
            " 1.6075035  1.61158375 1.59749312 1.61759682 1.61396952 1.63081581\n",
            " 1.60723898 1.60703001 1.61068619 1.62648783 1.60770815 1.61261958\n",
            " 0.70079511 1.607151   1.61169585 1.60296645 1.62020896 1.60844913\n",
            " 1.59892548 1.61339924 1.60671676 1.63017477 1.61561431 1.61174112\n",
            " 0.62420459 1.60703131 0.81259212 0.67819779 1.60916627 1.61326962\n",
            " 1.61081296 1.61481495 1.61839405 1.62126051 0.64806471 1.61022396\n",
            " 1.61477189 0.70626513 1.60769629 1.61260378 1.60290677 1.61351244\n",
            " 0.70678293 1.61292663 0.59211525 1.62166263 1.60159573 1.62194018\n",
            " 0.63238208 0.82123847 1.61195318 1.61410327 1.61355522 1.61704745\n",
            " 0.59438184 1.60583901 1.60966342 1.61341    0.70471087 1.6023314\n",
            " 1.62312087 1.60993122 1.60800106 1.62993369 1.6041506  1.6083538\n",
            " 1.60838651 1.59858842 1.60216128 0.77334694 0.70722799 1.61115044\n",
            " 1.61581666 0.69169723 1.60117251 0.7891823  1.61165307 1.61569487\n",
            " 1.61236578 1.61336467 0.6638307  1.60349581 1.6122279  1.60171316\n",
            " 1.61980551 1.6245243  0.60194886 1.6115707  0.79762305 1.61665588\n",
            " 1.61195282 1.61587097 1.61076407 1.60263689 1.62453528 0.66657317\n",
            " 1.61469055 1.61030913 1.61290311 1.60909696 1.61115896 1.62089533\n",
            " 1.61126227 1.61228902 1.60230632 1.61037774 0.82701979 0.78263152\n",
            " 0.69388244 0.80865367 0.7483892  1.62363556 0.79275674 1.61709981\n",
            " 0.52157477 0.70509036 0.73415834 1.61226502 0.74925864 0.69309126\n",
            " 0.70531862 0.64349008 0.81187463 1.60809465 1.61625862 1.61528032\n",
            " 1.62029648 0.80269209 0.69047622 1.6104549  1.61047362 1.6122932\n",
            " 0.66056689 0.61599189 1.60219965 0.71167104 1.61571376 0.67801698\n",
            " 0.63823522 1.61799442 1.60849798 1.60757432 0.69641649 1.610963\n",
            " 0.80671579 0.69741527 1.62620721 0.84103039 0.82084321 0.68928297\n",
            " 0.79586859 0.71390098 0.61078599 0.52049454 0.67551338 0.5989417\n",
            " 0.67163948 0.66971752 0.66817152 1.60799091 0.68854977 0.65850084\n",
            " 0.68216032 0.72841897 1.62740759 1.62153501 1.60820553 0.57357134\n",
            " 1.61179679 0.78616961 0.72174666 1.60639365 0.7909095  0.66847579\n",
            " 1.61349296 1.61717977 1.60767015 1.61227328 1.61070279 1.61629874\n",
            " 0.72449463 1.61334939 0.70728459 1.61517181 0.70353623 1.61444652\n",
            " 1.61485674]\n",
            "[tmin_K    0.985936\n",
            "tmax_K    0.635460\n",
            "dtype: float64, tmin_K    0.978321\n",
            "tmax_K    0.633645\n",
            "dtype: float64, tmin_K    0.972273\n",
            "tmax_K    0.632902\n",
            "dtype: float64, tmin_K    0.982850\n",
            "tmax_K    0.637208\n",
            "dtype: float64, tmin_K    0.980818\n",
            "tmax_K    0.636320\n",
            "dtype: float64, tmin_K    0.977420\n",
            "tmax_K    0.633492\n",
            "dtype: float64, tmin_K    0.972225\n",
            "tmax_K    0.632465\n",
            "dtype: float64, tmin_K    0.977667\n",
            "tmax_K    0.634004\n",
            "dtype: float64, tmin_K    0.977879\n",
            "tmax_K    0.631982\n",
            "dtype: float64, tmin_K    0.980284\n",
            "tmax_K    0.634369\n",
            "dtype: float64, tmin_K    0.980861\n",
            "tmax_K    0.636858\n",
            "dtype: float64, tmin_K    0.964608\n",
            "tmax_K    0.630473\n",
            "dtype: float64, tmin_K    0.975590\n",
            "tmax_K    0.630059\n",
            "dtype: float64, tmin_K    0.983733\n",
            "tmax_K    0.637679\n",
            "dtype: float64, tmin_K    0.978138\n",
            "tmax_K    0.631790\n",
            "dtype: float64, tmin_K    0.985373\n",
            "tmax_K    0.636630\n",
            "dtype: float64, tmin_K    0.984500\n",
            "tmax_K    0.635013\n",
            "dtype: float64, tmin_K    0.983826\n",
            "tmax_K    0.638332\n",
            "dtype: float64, tmin_K    0.977124\n",
            "tmax_K    0.630380\n",
            "dtype: float64, tmin_K    0.978319\n",
            "tmax_K    0.633265\n",
            "dtype: float64, tmin_K    0.969590\n",
            "tmax_K    0.627903\n",
            "dtype: float64, tmin_K    0.986040\n",
            "tmax_K    0.631556\n",
            "dtype: float64, tmin_K    0.980267\n",
            "tmax_K    0.633703\n",
            "dtype: float64, tmin_K    0.994150\n",
            "tmax_K    0.636666\n",
            "dtype: float64, tmin_K    0.974650\n",
            "tmax_K    0.632589\n",
            "dtype: float64, tmin_K    0.974478\n",
            "tmax_K    0.632552\n",
            "dtype: float64, tmin_K    0.976484\n",
            "tmax_K    0.634202\n",
            "dtype: float64, tmin_K    0.988475\n",
            "tmax_K    0.638013\n",
            "dtype: float64, tmin_K    0.975566\n",
            "tmax_K    0.632143\n",
            "dtype: float64, tmin_K    0.978796\n",
            "tmax_K    0.633823\n",
            "dtype: float64, tmin_K    0.350832\n",
            "tmax_K    0.349963\n",
            "dtype: float64, tmin_K    0.975594\n",
            "tmax_K    0.631557\n",
            "dtype: float64, tmin_K    0.977960\n",
            "tmax_K    0.633736\n",
            "dtype: float64, tmin_K    0.968848\n",
            "tmax_K    0.634118\n",
            "dtype: float64, tmin_K    0.985032\n",
            "tmax_K    0.635177\n",
            "dtype: float64, tmin_K    0.977362\n",
            "tmax_K    0.631087\n",
            "dtype: float64, tmin_K    0.972711\n",
            "tmax_K    0.626214\n",
            "dtype: float64, tmin_K    0.977995\n",
            "tmax_K    0.635404\n",
            "dtype: float64, tmin_K    0.975061\n",
            "tmax_K    0.631656\n",
            "dtype: float64, tmin_K    0.988590\n",
            "tmax_K    0.641585\n",
            "dtype: float64, tmin_K    0.980289\n",
            "tmax_K    0.635325\n",
            "dtype: float64, tmin_K    0.977494\n",
            "tmax_K    0.634248\n",
            "dtype: float64, tmin_K    0.319726\n",
            "tmax_K    0.304479\n",
            "dtype: float64, tmin_K    0.972603\n",
            "tmax_K    0.634428\n",
            "dtype: float64, tmin_K    0.468067\n",
            "tmax_K    0.344525\n",
            "dtype: float64, tmin_K    0.338264\n",
            "tmax_K    0.339934\n",
            "dtype: float64, tmin_K    0.977127\n",
            "tmax_K    0.632039\n",
            "dtype: float64, tmin_K    0.975184\n",
            "tmax_K    0.638086\n",
            "dtype: float64, tmin_K    0.977268\n",
            "tmax_K    0.633545\n",
            "dtype: float64, tmin_K    0.980263\n",
            "tmax_K    0.634552\n",
            "dtype: float64, tmin_K    0.982250\n",
            "tmax_K    0.636144\n",
            "dtype: float64, tmin_K    0.987948\n",
            "tmax_K    0.633313\n",
            "dtype: float64, tmin_K    0.338732\n",
            "tmax_K    0.309333\n",
            "dtype: float64, tmin_K    0.976781\n",
            "tmax_K    0.633443\n",
            "dtype: float64, tmin_K    0.980163\n",
            "tmax_K    0.634608\n",
            "dtype: float64, tmin_K    0.345709\n",
            "tmax_K    0.360556\n",
            "dtype: float64, tmin_K    0.974869\n",
            "tmax_K    0.632828\n",
            "dtype: float64, tmin_K    0.977450\n",
            "tmax_K    0.635153\n",
            "dtype: float64, tmin_K    0.970475\n",
            "tmax_K    0.632432\n",
            "dtype: float64, tmin_K    0.978313\n",
            "tmax_K    0.635199\n",
            "dtype: float64, tmin_K    0.342550\n",
            "tmax_K    0.364233\n",
            "dtype: float64, tmin_K    0.978882\n",
            "tmax_K    0.634045\n",
            "dtype: float64, tmin_K    0.311480\n",
            "tmax_K    0.280635\n",
            "dtype: float64, tmin_K    0.983749\n",
            "tmax_K    0.637913\n",
            "dtype: float64, tmin_K    0.973017\n",
            "tmax_K    0.628579\n",
            "dtype: float64, tmin_K    0.987021\n",
            "tmax_K    0.634920\n",
            "dtype: float64, tmin_K    0.329672\n",
            "tmax_K    0.302710\n",
            "dtype: float64, tmin_K    0.460957\n",
            "tmax_K    0.360281\n",
            "dtype: float64, tmin_K    0.977793\n",
            "tmax_K    0.634160\n",
            "dtype: float64, tmin_K    0.980597\n",
            "tmax_K    0.633506\n",
            "dtype: float64, tmin_K    0.978344\n",
            "tmax_K    0.635211\n",
            "dtype: float64, tmin_K    0.982347\n",
            "tmax_K    0.634701\n",
            "dtype: float64, tmin_K    0.301418\n",
            "tmax_K    0.292964\n",
            "dtype: float64, tmin_K    0.974115\n",
            "tmax_K    0.631724\n",
            "dtype: float64, tmin_K    0.975666\n",
            "tmax_K    0.633997\n",
            "dtype: float64, tmin_K    0.980036\n",
            "tmax_K    0.633374\n",
            "dtype: float64, tmin_K    0.408986\n",
            "tmax_K    0.295725\n",
            "dtype: float64, tmin_K    0.967869\n",
            "tmax_K    0.634463\n",
            "dtype: float64, tmin_K    0.987167\n",
            "tmax_K    0.635954\n",
            "dtype: float64, tmin_K    0.977567\n",
            "tmax_K    0.632364\n",
            "dtype: float64, tmin_K    0.976943\n",
            "tmax_K    0.631058\n",
            "dtype: float64, tmin_K    0.990245\n",
            "tmax_K    0.639689\n",
            "dtype: float64, tmin_K    0.971165\n",
            "tmax_K    0.632985\n",
            "dtype: float64, tmin_K    0.979039\n",
            "tmax_K    0.629314\n",
            "dtype: float64, tmin_K    0.975337\n",
            "tmax_K    0.633049\n",
            "dtype: float64, tmin_K    0.967386\n",
            "tmax_K    0.631203\n",
            "dtype: float64, tmin_K    0.972237\n",
            "tmax_K    0.629924\n",
            "dtype: float64, tmin_K    0.411403\n",
            "tmax_K    0.361944\n",
            "dtype: float64, tmin_K    0.359713\n",
            "tmax_K    0.347515\n",
            "dtype: float64, tmin_K    0.977339\n",
            "tmax_K    0.633811\n",
            "dtype: float64, tmin_K    0.976947\n",
            "tmax_K    0.638870\n",
            "dtype: float64, tmin_K    0.372536\n",
            "tmax_K    0.319162\n",
            "dtype: float64, tmin_K    0.966826\n",
            "tmax_K    0.634347\n",
            "dtype: float64, tmin_K    0.436729\n",
            "tmax_K    0.352454\n",
            "dtype: float64, tmin_K    0.977487\n",
            "tmax_K    0.634166\n",
            "dtype: float64, tmin_K    0.981864\n",
            "tmax_K    0.633831\n",
            "dtype: float64, tmin_K    0.978381\n",
            "tmax_K    0.633985\n",
            "dtype: float64, tmin_K    0.979036\n",
            "tmax_K    0.634328\n",
            "dtype: float64, tmin_K    0.346998\n",
            "tmax_K    0.316833\n",
            "dtype: float64, tmin_K    0.970974\n",
            "tmax_K    0.632522\n",
            "dtype: float64, tmin_K    0.977956\n",
            "tmax_K    0.634272\n",
            "dtype: float64, tmin_K    0.970138\n",
            "tmax_K    0.631575\n",
            "dtype: float64, tmin_K    0.988762\n",
            "tmax_K    0.631044\n",
            "dtype: float64, tmin_K    0.990534\n",
            "tmax_K    0.633991\n",
            "dtype: float64, tmin_K    0.326670\n",
            "tmax_K    0.275278\n",
            "dtype: float64, tmin_K    0.978062\n",
            "tmax_K    0.633508\n",
            "dtype: float64, tmin_K    0.435321\n",
            "tmax_K    0.362302\n",
            "dtype: float64, tmin_K    0.984536\n",
            "tmax_K    0.632120\n",
            "dtype: float64, tmin_K    0.976186\n",
            "tmax_K    0.635767\n",
            "dtype: float64, tmin_K    0.980900\n",
            "tmax_K    0.634971\n",
            "dtype: float64, tmin_K    0.977507\n",
            "tmax_K    0.633257\n",
            "dtype: float64, tmin_K    0.970511\n",
            "tmax_K    0.632126\n",
            "dtype: float64, tmin_K    0.986632\n",
            "tmax_K    0.637903\n",
            "dtype: float64, tmin_K    0.350028\n",
            "tmax_K    0.316545\n",
            "dtype: float64, tmin_K    0.977561\n",
            "tmax_K    0.637130\n",
            "dtype: float64, tmin_K    0.977237\n",
            "tmax_K    0.633073\n",
            "dtype: float64, tmin_K    0.978519\n",
            "tmax_K    0.634384\n",
            "dtype: float64, tmin_K    0.976209\n",
            "tmax_K    0.632887\n",
            "dtype: float64, tmin_K    0.978424\n",
            "tmax_K    0.632735\n",
            "dtype: float64, tmin_K    0.982859\n",
            "tmax_K    0.638037\n",
            "dtype: float64, tmin_K    0.976910\n",
            "tmax_K    0.634352\n",
            "dtype: float64, tmin_K    0.978688\n",
            "tmax_K    0.633601\n",
            "dtype: float64, tmin_K    0.971325\n",
            "tmax_K    0.630981\n",
            "dtype: float64, tmin_K    0.977178\n",
            "tmax_K    0.633200\n",
            "dtype: float64, tmin_K    0.463280\n",
            "tmax_K    0.363739\n",
            "dtype: float64, tmin_K    0.436675\n",
            "tmax_K    0.345956\n",
            "dtype: float64, tmin_K    0.380981\n",
            "tmax_K    0.312902\n",
            "dtype: float64, tmin_K    0.449606\n",
            "tmax_K    0.359048\n",
            "dtype: float64, tmin_K    0.423237\n",
            "tmax_K    0.325153\n",
            "dtype: float64, tmin_K    0.983799\n",
            "tmax_K    0.639836\n",
            "dtype: float64, tmin_K    0.449419\n",
            "tmax_K    0.343337\n",
            "dtype: float64, tmin_K    0.982297\n",
            "tmax_K    0.634803\n",
            "dtype: float64, tmin_K    0.277394\n",
            "tmax_K    0.244181\n",
            "dtype: float64, tmin_K    0.392451\n",
            "tmax_K    0.312640\n",
            "dtype: float64, tmin_K    0.435825\n",
            "tmax_K    0.298333\n",
            "dtype: float64, tmin_K    0.978210\n",
            "tmax_K    0.634055\n",
            "dtype: float64, tmin_K    0.424146\n",
            "tmax_K    0.325113\n",
            "dtype: float64, tmin_K    0.337340\n",
            "tmax_K    0.355751\n",
            "dtype: float64, tmin_K    0.363606\n",
            "tmax_K    0.341713\n",
            "dtype: float64, tmin_K    0.358494\n",
            "tmax_K    0.284996\n",
            "dtype: float64, tmin_K    0.467204\n",
            "tmax_K    0.344671\n",
            "dtype: float64, tmin_K    0.975636\n",
            "tmax_K    0.632459\n",
            "dtype: float64, tmin_K    0.985993\n",
            "tmax_K    0.630265\n",
            "dtype: float64, tmin_K    0.980392\n",
            "tmax_K    0.634889\n",
            "dtype: float64, tmin_K    0.982248\n",
            "tmax_K    0.638048\n",
            "dtype: float64, tmin_K    0.454718\n",
            "tmax_K    0.347975\n",
            "dtype: float64, tmin_K    0.323566\n",
            "tmax_K    0.366910\n",
            "dtype: float64, tmin_K    0.976764\n",
            "tmax_K    0.633691\n",
            "dtype: float64, tmin_K    0.977178\n",
            "tmax_K    0.633296\n",
            "dtype: float64, tmin_K    0.977973\n",
            "tmax_K    0.634320\n",
            "dtype: float64, tmin_K    0.393268\n",
            "tmax_K    0.267299\n",
            "dtype: float64, tmin_K    0.344033\n",
            "tmax_K    0.271959\n",
            "dtype: float64, tmin_K    0.976173\n",
            "tmax_K    0.626027\n",
            "dtype: float64, tmin_K    0.345469\n",
            "tmax_K    0.366202\n",
            "dtype: float64, tmin_K    0.981843\n",
            "tmax_K    0.633871\n",
            "dtype: float64, tmin_K    0.335524\n",
            "tmax_K    0.342493\n",
            "dtype: float64, tmin_K    0.291814\n",
            "tmax_K    0.346421\n",
            "dtype: float64, tmin_K    0.982469\n",
            "tmax_K    0.635525\n",
            "dtype: float64, tmin_K    0.973704\n",
            "tmax_K    0.634794\n",
            "dtype: float64, tmin_K    0.977728\n",
            "tmax_K    0.629846\n",
            "dtype: float64, tmin_K    0.335358\n",
            "tmax_K    0.361059\n",
            "dtype: float64, tmin_K    0.976734\n",
            "tmax_K    0.634229\n",
            "dtype: float64, tmin_K    0.446506\n",
            "tmax_K    0.360210\n",
            "dtype: float64, tmin_K    0.373723\n",
            "tmax_K    0.323692\n",
            "dtype: float64, tmin_K    1.015269\n",
            "tmax_K    0.610938\n",
            "dtype: float64, tmin_K    0.460593\n",
            "tmax_K    0.380437\n",
            "dtype: float64, tmin_K    0.469710\n",
            "tmax_K    0.351133\n",
            "dtype: float64, tmin_K    0.370489\n",
            "tmax_K    0.318794\n",
            "dtype: float64, tmin_K    0.445201\n",
            "tmax_K    0.350667\n",
            "dtype: float64, tmin_K    0.342178\n",
            "tmax_K    0.371723\n",
            "dtype: float64, tmin_K    0.319326\n",
            "tmax_K    0.291460\n",
            "dtype: float64, tmin_K    0.268922\n",
            "tmax_K    0.251573\n",
            "dtype: float64, tmin_K    0.343855\n",
            "tmax_K    0.331658\n",
            "dtype: float64, tmin_K    0.368756\n",
            "tmax_K    0.230186\n",
            "dtype: float64, tmin_K    0.346131\n",
            "tmax_K    0.325509\n",
            "dtype: float64, tmin_K    0.389462\n",
            "tmax_K    0.280255\n",
            "dtype: float64, tmin_K    0.354533\n",
            "tmax_K    0.313638\n",
            "dtype: float64, tmin_K    0.977042\n",
            "tmax_K    0.630949\n",
            "dtype: float64, tmin_K    0.364255\n",
            "tmax_K    0.324294\n",
            "dtype: float64, tmin_K    0.342781\n",
            "tmax_K    0.315720\n",
            "dtype: float64, tmin_K    0.339056\n",
            "tmax_K    0.343104\n",
            "dtype: float64, tmin_K    0.378877\n",
            "tmax_K    0.349542\n",
            "dtype: float64, tmin_K    0.988280\n",
            "tmax_K    0.639127\n",
            "dtype: float64, tmin_K    0.985424\n",
            "tmax_K    0.636111\n",
            "dtype: float64, tmin_K    0.973728\n",
            "tmax_K    0.634477\n",
            "dtype: float64, tmin_K    0.313441\n",
            "tmax_K    0.260130\n",
            "dtype: float64, tmin_K    0.979336\n",
            "tmax_K    0.632461\n",
            "dtype: float64, tmin_K    0.435215\n",
            "tmax_K    0.350955\n",
            "dtype: float64, tmin_K    0.363963\n",
            "tmax_K    0.357784\n",
            "dtype: float64, tmin_K    0.972099\n",
            "tmax_K    0.634295\n",
            "dtype: float64, tmin_K    0.43363\n",
            "tmax_K    0.35728\n",
            "dtype: float64, tmin_K    0.368634\n",
            "tmax_K    0.299841\n",
            "dtype: float64, tmin_K    0.980416\n",
            "tmax_K    0.633077\n",
            "dtype: float64, tmin_K    0.982997\n",
            "tmax_K    0.634182\n",
            "dtype: float64, tmin_K    0.974954\n",
            "tmax_K    0.632716\n",
            "dtype: float64, tmin_K    0.977270\n",
            "tmax_K    0.635004\n",
            "dtype: float64, tmin_K    0.979443\n",
            "tmax_K    0.631260\n",
            "dtype: float64, tmin_K    0.975783\n",
            "tmax_K    0.640516\n",
            "dtype: float64, tmin_K    0.375129\n",
            "tmax_K    0.349366\n",
            "dtype: float64, tmin_K    0.979046\n",
            "tmax_K    0.634303\n",
            "dtype: float64, tmin_K    0.363352\n",
            "tmax_K    0.343933\n",
            "dtype: float64, tmin_K    0.978003\n",
            "tmax_K    0.637168\n",
            "dtype: float64, tmin_K    0.395366\n",
            "tmax_K    0.308170\n",
            "dtype: float64, tmin_K    0.978732\n",
            "tmax_K    0.635714\n",
            "dtype: float64, tmin_K    0.978648\n",
            "tmax_K    0.636209\n",
            "dtype: float64]\n",
            "\t BEST: 5 hidden layers, 17 hidden neurons\n",
            "Model Number: 28\n",
            "\t # Hidden Layers = 1\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.977263\n",
            "tmax_K    0.635805\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.966669\n",
            "tmax_K    0.628808\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.976774\n",
            "tmax_K    0.634415\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.985091\n",
            "tmax_K    0.636390\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.980074\n",
            "tmax_K    0.635529\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.980697\n",
            "tmax_K    0.633577\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.981273\n",
            "tmax_K    0.634084\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.975996\n",
            "tmax_K    0.633722\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.965678\n",
            "tmax_K    0.631153\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.970321\n",
            "tmax_K    0.633118\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.973132\n",
            "tmax_K    0.632453\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.990382\n",
            "tmax_K    0.636574\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.971974\n",
            "tmax_K    0.632403\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.973514\n",
            "tmax_K    0.631119\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.966641\n",
            "tmax_K    0.632008\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.944724\n",
            "tmax_K    0.632756\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.971698\n",
            "tmax_K    0.637765\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.974589\n",
            "tmax_K    0.632056\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.981380\n",
            "tmax_K    0.635286\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.978951\n",
            "tmax_K    0.634669\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.984028\n",
            "tmax_K    0.637743\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.985526\n",
            "tmax_K    0.632217\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.983734\n",
            "tmax_K    0.637399\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.963983\n",
            "tmax_K    0.627016\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.339986\n",
            "tmax_K    0.321423\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.975477\n",
            "tmax_K    0.632720\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.967253\n",
            "tmax_K    0.639896\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.999734\n",
            "tmax_K    0.642701\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.982369\n",
            "tmax_K    0.635916\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.985465\n",
            "tmax_K    0.634052\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.978777\n",
            "tmax_K    0.638424\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.982382\n",
            "tmax_K    0.634885\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.976650\n",
            "tmax_K    0.632478\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.982695\n",
            "tmax_K    0.633801\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.981977\n",
            "tmax_K    0.635356\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.969668\n",
            "tmax_K    0.630728\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.977171\n",
            "tmax_K    0.630456\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.972578\n",
            "tmax_K    0.626484\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.988816\n",
            "tmax_K    0.634016\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.978972\n",
            "tmax_K    0.632853\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.319301\n",
            "tmax_K    0.342660\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 2\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.986211\n",
            "tmax_K    0.636339\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.976940\n",
            "tmax_K    0.635542\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.455169\n",
            "tmax_K    0.367960\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.977969\n",
            "tmax_K    0.634613\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.368498\n",
            "tmax_K    0.347950\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.985431\n",
            "tmax_K    0.633676\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.395802\n",
            "tmax_K    0.379863\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.982147\n",
            "tmax_K    0.636504\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.976936\n",
            "tmax_K    0.632707\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.979753\n",
            "tmax_K    0.634546\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.966871\n",
            "tmax_K    0.629993\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.975774\n",
            "tmax_K    0.637894\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.299774\n",
            "tmax_K    0.357572\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.983177\n",
            "tmax_K    0.636072\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.447238\n",
            "tmax_K    0.367468\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.978820\n",
            "tmax_K    0.633408\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.978946\n",
            "tmax_K    0.633693\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.982427\n",
            "tmax_K    0.637667\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.980587\n",
            "tmax_K    0.635076\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.980476\n",
            "tmax_K    0.634751\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.393779\n",
            "tmax_K    0.309449\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.980215\n",
            "tmax_K    0.633516\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.97952\n",
            "tmax_K    0.63434\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.983795\n",
            "tmax_K    0.635900\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.983426\n",
            "tmax_K    0.635157\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.983218\n",
            "tmax_K    0.636077\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.978770\n",
            "tmax_K    0.632493\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.982835\n",
            "tmax_K    0.633159\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.980601\n",
            "tmax_K    0.635217\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.972971\n",
            "tmax_K    0.630589\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.981354\n",
            "tmax_K    0.634825\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.981691\n",
            "tmax_K    0.634651\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.974860\n",
            "tmax_K    0.633384\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.974610\n",
            "tmax_K    0.631672\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.977209\n",
            "tmax_K    0.629906\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.98541\n",
            "tmax_K    0.63598\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.970499\n",
            "tmax_K    0.630615\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.977414\n",
            "tmax_K    0.637114\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.992438\n",
            "tmax_K    0.633903\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.983944\n",
            "tmax_K    0.635825\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.975636\n",
            "tmax_K    0.630301\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 3\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.466749\n",
            "tmax_K    0.343916\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.436771\n",
            "tmax_K    0.380640\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.980012\n",
            "tmax_K    0.637145\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.423095\n",
            "tmax_K    0.337566\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.975826\n",
            "tmax_K    0.634228\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.978397\n",
            "tmax_K    0.634245\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.986778\n",
            "tmax_K    0.632454\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.980146\n",
            "tmax_K    0.634869\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.981902\n",
            "tmax_K    0.634771\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.981271\n",
            "tmax_K    0.634124\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.398630\n",
            "tmax_K    0.300686\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.369860\n",
            "tmax_K    0.336906\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.977856\n",
            "tmax_K    0.634109\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.977293\n",
            "tmax_K    0.634066\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.980585\n",
            "tmax_K    0.640353\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.979402\n",
            "tmax_K    0.633873\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.977417\n",
            "tmax_K    0.632808\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.979015\n",
            "tmax_K    0.635077\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.978669\n",
            "tmax_K    0.639013\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.979315\n",
            "tmax_K    0.633258\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.977397\n",
            "tmax_K    0.633713\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.979414\n",
            "tmax_K    0.633602\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.962180\n",
            "tmax_K    0.625923\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.974608\n",
            "tmax_K    0.631232\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.445380\n",
            "tmax_K    0.356872\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.980648\n",
            "tmax_K    0.633092\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.317778\n",
            "tmax_K    0.322723\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.978149\n",
            "tmax_K    0.630740\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.371099\n",
            "tmax_K    0.315002\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.979465\n",
            "tmax_K    0.634733\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.360164\n",
            "tmax_K    0.359987\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.976298\n",
            "tmax_K    0.632161\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.974098\n",
            "tmax_K    0.630348\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.983621\n",
            "tmax_K    0.620816\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.975433\n",
            "tmax_K    0.632638\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.339854\n",
            "tmax_K    0.350268\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.980539\n",
            "tmax_K    0.634059\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.980044\n",
            "tmax_K    0.634261\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.979682\n",
            "tmax_K    0.636347\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.975683\n",
            "tmax_K    0.630959\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.986291\n",
            "tmax_K    0.637282\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 4\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.430143\n",
            "tmax_K    0.342735\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.970686\n",
            "tmax_K    0.633246\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.466063\n",
            "tmax_K    0.361931\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.974769\n",
            "tmax_K    0.632569\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.978011\n",
            "tmax_K    0.627626\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.382851\n",
            "tmax_K    0.308338\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.979088\n",
            "tmax_K    0.634931\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.305968\n",
            "tmax_K    0.326398\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.987075\n",
            "tmax_K    0.635218\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.974502\n",
            "tmax_K    0.631047\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.976507\n",
            "tmax_K    0.633482\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.978060\n",
            "tmax_K    0.634132\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.977007\n",
            "tmax_K    0.633531\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.461137\n",
            "tmax_K    0.371716\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.454210\n",
            "tmax_K    0.367249\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.340547\n",
            "tmax_K    0.299952\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.395491\n",
            "tmax_K    0.302373\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.980705\n",
            "tmax_K    0.635496\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.338351\n",
            "tmax_K    0.359649\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.437244\n",
            "tmax_K    0.374736\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.988813\n",
            "tmax_K    0.631663\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.36297\n",
            "tmax_K    0.34031\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.357387\n",
            "tmax_K    0.264232\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.308316\n",
            "tmax_K    0.305389\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.981792\n",
            "tmax_K    0.626053\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.975541\n",
            "tmax_K    0.632568\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.968316\n",
            "tmax_K    0.634597\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.467982\n",
            "tmax_K    0.356662\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.975285\n",
            "tmax_K    0.630996\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.978419\n",
            "tmax_K    0.636193\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.981497\n",
            "tmax_K    0.636800\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.313749\n",
            "tmax_K    0.306200\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.982689\n",
            "tmax_K    0.638123\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.330252\n",
            "tmax_K    0.312692\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.981521\n",
            "tmax_K    0.635903\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.353547\n",
            "tmax_K    0.336769\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.976212\n",
            "tmax_K    0.632195\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.422608\n",
            "tmax_K    0.376753\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.973975\n",
            "tmax_K    0.633220\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.976044\n",
            "tmax_K    0.633949\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.363860\n",
            "tmax_K    0.327365\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 5\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.453949\n",
            "tmax_K    0.341606\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.449320\n",
            "tmax_K    0.359887\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.456126\n",
            "tmax_K    0.350853\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.455106\n",
            "tmax_K    0.324437\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.365982\n",
            "tmax_K    0.339302\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.438270\n",
            "tmax_K    0.362841\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.429728\n",
            "tmax_K    0.343519\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.353281\n",
            "tmax_K    0.313478\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.323391\n",
            "tmax_K    0.354663\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.982153\n",
            "tmax_K    0.636035\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.975825\n",
            "tmax_K    0.633433\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.425444\n",
            "tmax_K    0.319103\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.979445\n",
            "tmax_K    0.634201\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.971717\n",
            "tmax_K    0.631886\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.978361\n",
            "tmax_K    0.632329\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.976125\n",
            "tmax_K    0.639254\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.440360\n",
            "tmax_K    0.359743\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.368046\n",
            "tmax_K    0.305932\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.357575\n",
            "tmax_K    0.347460\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.388369\n",
            "tmax_K    0.326634\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.979147\n",
            "tmax_K    0.633396\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.978733\n",
            "tmax_K    0.634972\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.447676\n",
            "tmax_K    0.323205\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.289707\n",
            "tmax_K    0.378911\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.328734\n",
            "tmax_K    0.278747\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.975578\n",
            "tmax_K    0.630399\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.980854\n",
            "tmax_K    0.635246\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.974590\n",
            "tmax_K    0.631754\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.982663\n",
            "tmax_K    0.636759\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.333881\n",
            "tmax_K    0.357414\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.307684\n",
            "tmax_K    0.304865\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.338079\n",
            "tmax_K    0.363604\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.374191\n",
            "tmax_K    0.382758\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.326762\n",
            "tmax_K    0.333948\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.986447\n",
            "tmax_K    0.640421\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.977029\n",
            "tmax_K    0.633420\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.976859\n",
            "tmax_K    0.632719\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.341597\n",
            "tmax_K    0.285552\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.979394\n",
            "tmax_K    0.633072\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.979499\n",
            "tmax_K    0.633337\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.310325\n",
            "tmax_K    0.289266\n",
            "dtype: float64\n",
            "[1.61306826 1.59547761 1.61118911 1.62148098 1.61560378 1.61427357\n",
            " 1.61535646 1.60971809 1.59683112 1.60343945 1.60558449 1.62695602\n",
            " 1.60437652 1.6046334  1.59864932 1.57747954 1.60946239 1.60664585\n",
            " 1.61666568 1.61362009 1.6217704  1.61774315 1.6211331  1.59099889\n",
            " 0.66140931 1.60819744 1.60714867 1.64243557 1.61828535 1.61951664\n",
            " 1.61720102 1.61726687 1.60912778 1.61649527 1.61733339 1.60039642\n",
            " 1.6076274  1.59906145 1.62283214 1.61182509 0.66196057 1.62254981\n",
            " 1.61248158 0.82312896 1.61258192 0.71644834 1.619107   0.77566464\n",
            " 1.61865089 1.60964341 1.61429929 1.5968643  1.61366824 0.65734678\n",
            " 1.61924926 0.81470601 1.61222748 1.61263833 1.62009396 1.61566331\n",
            " 1.61522653 0.70322834 1.61373098 1.61385971 1.6196949  1.61858386\n",
            " 1.61929494 1.6112631  1.61599406 1.61581871 1.60356014 1.61617953\n",
            " 1.61634271 1.60824366 1.60628268 1.60711496 1.6213894  1.60111445\n",
            " 1.61452872 1.62634082 1.61976888 1.60593669 0.81066483 0.81741069\n",
            " 1.61715645 0.76066096 1.61005417 1.61264199 1.6192322  1.61501496\n",
            " 1.61667337 1.61539494 0.69931629 0.70676677 1.61196471 1.61135926\n",
            " 1.62093814 1.61327534 1.61022515 1.61409128 1.61768163 1.61257291\n",
            " 1.61111039 1.61301609 1.58810268 1.60584026 0.80225217 1.61373993\n",
            " 0.64050103 1.6088884  0.68610169 1.61419855 0.72015127 1.6084594\n",
            " 1.60444607 1.60443708 1.60807076 0.69012148 1.61459799 1.61430521\n",
            " 1.61602886 1.60664211 1.62357318 0.77287776 1.60393236 0.82799485\n",
            " 1.60733845 1.60563689 0.69118814 1.61401913 0.63236554 1.62229352\n",
            " 1.60554882 1.6099887  1.61219252 1.61053737 0.83285338 0.821459\n",
            " 0.64049882 0.69786406 1.61620048 0.69800044 0.81197982 1.62047625\n",
            " 0.70327977 0.62161877 0.61370512 1.60784489 1.60810831 1.60291236\n",
            " 0.82464336 1.60628071 1.61461213 1.61829746 0.61994863 1.62081242\n",
            " 0.64294353 1.61742394 0.6903158  1.60840701 0.79936128 1.60719476\n",
            " 1.60999329 0.69122471 0.7955559  0.80920733 0.80697833 0.77954313\n",
            " 0.70528389 0.80111019 0.77324752 0.66675862 0.67805416 1.61818795\n",
            " 1.60925775 0.74454669 1.61364549 1.60360273 1.61069082 1.61537932\n",
            " 0.80010318 0.67397793 0.70503542 0.71500286 1.61254288 1.6137056\n",
            " 0.77088082 0.66861889 0.60748071 1.60597683 1.61609981 1.60634497\n",
            " 1.61942255 0.69129417 0.61254881 0.70168379 0.75694889 0.66071015\n",
            " 1.62686813 1.61044865 1.60957815 0.62714911 1.61246565 1.61283563\n",
            " 0.5995905 ]\n",
            "[tmin_K    0.977263\n",
            "tmax_K    0.635805\n",
            "dtype: float64, tmin_K    0.966669\n",
            "tmax_K    0.628808\n",
            "dtype: float64, tmin_K    0.976774\n",
            "tmax_K    0.634415\n",
            "dtype: float64, tmin_K    0.985091\n",
            "tmax_K    0.636390\n",
            "dtype: float64, tmin_K    0.980074\n",
            "tmax_K    0.635529\n",
            "dtype: float64, tmin_K    0.980697\n",
            "tmax_K    0.633577\n",
            "dtype: float64, tmin_K    0.981273\n",
            "tmax_K    0.634084\n",
            "dtype: float64, tmin_K    0.975996\n",
            "tmax_K    0.633722\n",
            "dtype: float64, tmin_K    0.965678\n",
            "tmax_K    0.631153\n",
            "dtype: float64, tmin_K    0.970321\n",
            "tmax_K    0.633118\n",
            "dtype: float64, tmin_K    0.973132\n",
            "tmax_K    0.632453\n",
            "dtype: float64, tmin_K    0.990382\n",
            "tmax_K    0.636574\n",
            "dtype: float64, tmin_K    0.971974\n",
            "tmax_K    0.632403\n",
            "dtype: float64, tmin_K    0.973514\n",
            "tmax_K    0.631119\n",
            "dtype: float64, tmin_K    0.966641\n",
            "tmax_K    0.632008\n",
            "dtype: float64, tmin_K    0.944724\n",
            "tmax_K    0.632756\n",
            "dtype: float64, tmin_K    0.971698\n",
            "tmax_K    0.637765\n",
            "dtype: float64, tmin_K    0.974589\n",
            "tmax_K    0.632056\n",
            "dtype: float64, tmin_K    0.981380\n",
            "tmax_K    0.635286\n",
            "dtype: float64, tmin_K    0.978951\n",
            "tmax_K    0.634669\n",
            "dtype: float64, tmin_K    0.984028\n",
            "tmax_K    0.637743\n",
            "dtype: float64, tmin_K    0.985526\n",
            "tmax_K    0.632217\n",
            "dtype: float64, tmin_K    0.983734\n",
            "tmax_K    0.637399\n",
            "dtype: float64, tmin_K    0.963983\n",
            "tmax_K    0.627016\n",
            "dtype: float64, tmin_K    0.339986\n",
            "tmax_K    0.321423\n",
            "dtype: float64, tmin_K    0.975477\n",
            "tmax_K    0.632720\n",
            "dtype: float64, tmin_K    0.967253\n",
            "tmax_K    0.639896\n",
            "dtype: float64, tmin_K    0.999734\n",
            "tmax_K    0.642701\n",
            "dtype: float64, tmin_K    0.982369\n",
            "tmax_K    0.635916\n",
            "dtype: float64, tmin_K    0.985465\n",
            "tmax_K    0.634052\n",
            "dtype: float64, tmin_K    0.978777\n",
            "tmax_K    0.638424\n",
            "dtype: float64, tmin_K    0.982382\n",
            "tmax_K    0.634885\n",
            "dtype: float64, tmin_K    0.976650\n",
            "tmax_K    0.632478\n",
            "dtype: float64, tmin_K    0.982695\n",
            "tmax_K    0.633801\n",
            "dtype: float64, tmin_K    0.981977\n",
            "tmax_K    0.635356\n",
            "dtype: float64, tmin_K    0.969668\n",
            "tmax_K    0.630728\n",
            "dtype: float64, tmin_K    0.977171\n",
            "tmax_K    0.630456\n",
            "dtype: float64, tmin_K    0.972578\n",
            "tmax_K    0.626484\n",
            "dtype: float64, tmin_K    0.988816\n",
            "tmax_K    0.634016\n",
            "dtype: float64, tmin_K    0.978972\n",
            "tmax_K    0.632853\n",
            "dtype: float64, tmin_K    0.319301\n",
            "tmax_K    0.342660\n",
            "dtype: float64, tmin_K    0.986211\n",
            "tmax_K    0.636339\n",
            "dtype: float64, tmin_K    0.976940\n",
            "tmax_K    0.635542\n",
            "dtype: float64, tmin_K    0.455169\n",
            "tmax_K    0.367960\n",
            "dtype: float64, tmin_K    0.977969\n",
            "tmax_K    0.634613\n",
            "dtype: float64, tmin_K    0.368498\n",
            "tmax_K    0.347950\n",
            "dtype: float64, tmin_K    0.985431\n",
            "tmax_K    0.633676\n",
            "dtype: float64, tmin_K    0.395802\n",
            "tmax_K    0.379863\n",
            "dtype: float64, tmin_K    0.982147\n",
            "tmax_K    0.636504\n",
            "dtype: float64, tmin_K    0.976936\n",
            "tmax_K    0.632707\n",
            "dtype: float64, tmin_K    0.979753\n",
            "tmax_K    0.634546\n",
            "dtype: float64, tmin_K    0.966871\n",
            "tmax_K    0.629993\n",
            "dtype: float64, tmin_K    0.975774\n",
            "tmax_K    0.637894\n",
            "dtype: float64, tmin_K    0.299774\n",
            "tmax_K    0.357572\n",
            "dtype: float64, tmin_K    0.983177\n",
            "tmax_K    0.636072\n",
            "dtype: float64, tmin_K    0.447238\n",
            "tmax_K    0.367468\n",
            "dtype: float64, tmin_K    0.978820\n",
            "tmax_K    0.633408\n",
            "dtype: float64, tmin_K    0.978946\n",
            "tmax_K    0.633693\n",
            "dtype: float64, tmin_K    0.982427\n",
            "tmax_K    0.637667\n",
            "dtype: float64, tmin_K    0.980587\n",
            "tmax_K    0.635076\n",
            "dtype: float64, tmin_K    0.980476\n",
            "tmax_K    0.634751\n",
            "dtype: float64, tmin_K    0.393779\n",
            "tmax_K    0.309449\n",
            "dtype: float64, tmin_K    0.980215\n",
            "tmax_K    0.633516\n",
            "dtype: float64, tmin_K    0.97952\n",
            "tmax_K    0.63434\n",
            "dtype: float64, tmin_K    0.983795\n",
            "tmax_K    0.635900\n",
            "dtype: float64, tmin_K    0.983426\n",
            "tmax_K    0.635157\n",
            "dtype: float64, tmin_K    0.983218\n",
            "tmax_K    0.636077\n",
            "dtype: float64, tmin_K    0.978770\n",
            "tmax_K    0.632493\n",
            "dtype: float64, tmin_K    0.982835\n",
            "tmax_K    0.633159\n",
            "dtype: float64, tmin_K    0.980601\n",
            "tmax_K    0.635217\n",
            "dtype: float64, tmin_K    0.972971\n",
            "tmax_K    0.630589\n",
            "dtype: float64, tmin_K    0.981354\n",
            "tmax_K    0.634825\n",
            "dtype: float64, tmin_K    0.981691\n",
            "tmax_K    0.634651\n",
            "dtype: float64, tmin_K    0.974860\n",
            "tmax_K    0.633384\n",
            "dtype: float64, tmin_K    0.974610\n",
            "tmax_K    0.631672\n",
            "dtype: float64, tmin_K    0.977209\n",
            "tmax_K    0.629906\n",
            "dtype: float64, tmin_K    0.98541\n",
            "tmax_K    0.63598\n",
            "dtype: float64, tmin_K    0.970499\n",
            "tmax_K    0.630615\n",
            "dtype: float64, tmin_K    0.977414\n",
            "tmax_K    0.637114\n",
            "dtype: float64, tmin_K    0.992438\n",
            "tmax_K    0.633903\n",
            "dtype: float64, tmin_K    0.983944\n",
            "tmax_K    0.635825\n",
            "dtype: float64, tmin_K    0.975636\n",
            "tmax_K    0.630301\n",
            "dtype: float64, tmin_K    0.466749\n",
            "tmax_K    0.343916\n",
            "dtype: float64, tmin_K    0.436771\n",
            "tmax_K    0.380640\n",
            "dtype: float64, tmin_K    0.980012\n",
            "tmax_K    0.637145\n",
            "dtype: float64, tmin_K    0.423095\n",
            "tmax_K    0.337566\n",
            "dtype: float64, tmin_K    0.975826\n",
            "tmax_K    0.634228\n",
            "dtype: float64, tmin_K    0.978397\n",
            "tmax_K    0.634245\n",
            "dtype: float64, tmin_K    0.986778\n",
            "tmax_K    0.632454\n",
            "dtype: float64, tmin_K    0.980146\n",
            "tmax_K    0.634869\n",
            "dtype: float64, tmin_K    0.981902\n",
            "tmax_K    0.634771\n",
            "dtype: float64, tmin_K    0.981271\n",
            "tmax_K    0.634124\n",
            "dtype: float64, tmin_K    0.398630\n",
            "tmax_K    0.300686\n",
            "dtype: float64, tmin_K    0.369860\n",
            "tmax_K    0.336906\n",
            "dtype: float64, tmin_K    0.977856\n",
            "tmax_K    0.634109\n",
            "dtype: float64, tmin_K    0.977293\n",
            "tmax_K    0.634066\n",
            "dtype: float64, tmin_K    0.980585\n",
            "tmax_K    0.640353\n",
            "dtype: float64, tmin_K    0.979402\n",
            "tmax_K    0.633873\n",
            "dtype: float64, tmin_K    0.977417\n",
            "tmax_K    0.632808\n",
            "dtype: float64, tmin_K    0.979015\n",
            "tmax_K    0.635077\n",
            "dtype: float64, tmin_K    0.978669\n",
            "tmax_K    0.639013\n",
            "dtype: float64, tmin_K    0.979315\n",
            "tmax_K    0.633258\n",
            "dtype: float64, tmin_K    0.977397\n",
            "tmax_K    0.633713\n",
            "dtype: float64, tmin_K    0.979414\n",
            "tmax_K    0.633602\n",
            "dtype: float64, tmin_K    0.962180\n",
            "tmax_K    0.625923\n",
            "dtype: float64, tmin_K    0.974608\n",
            "tmax_K    0.631232\n",
            "dtype: float64, tmin_K    0.445380\n",
            "tmax_K    0.356872\n",
            "dtype: float64, tmin_K    0.980648\n",
            "tmax_K    0.633092\n",
            "dtype: float64, tmin_K    0.317778\n",
            "tmax_K    0.322723\n",
            "dtype: float64, tmin_K    0.978149\n",
            "tmax_K    0.630740\n",
            "dtype: float64, tmin_K    0.371099\n",
            "tmax_K    0.315002\n",
            "dtype: float64, tmin_K    0.979465\n",
            "tmax_K    0.634733\n",
            "dtype: float64, tmin_K    0.360164\n",
            "tmax_K    0.359987\n",
            "dtype: float64, tmin_K    0.976298\n",
            "tmax_K    0.632161\n",
            "dtype: float64, tmin_K    0.974098\n",
            "tmax_K    0.630348\n",
            "dtype: float64, tmin_K    0.983621\n",
            "tmax_K    0.620816\n",
            "dtype: float64, tmin_K    0.975433\n",
            "tmax_K    0.632638\n",
            "dtype: float64, tmin_K    0.339854\n",
            "tmax_K    0.350268\n",
            "dtype: float64, tmin_K    0.980539\n",
            "tmax_K    0.634059\n",
            "dtype: float64, tmin_K    0.980044\n",
            "tmax_K    0.634261\n",
            "dtype: float64, tmin_K    0.979682\n",
            "tmax_K    0.636347\n",
            "dtype: float64, tmin_K    0.975683\n",
            "tmax_K    0.630959\n",
            "dtype: float64, tmin_K    0.986291\n",
            "tmax_K    0.637282\n",
            "dtype: float64, tmin_K    0.430143\n",
            "tmax_K    0.342735\n",
            "dtype: float64, tmin_K    0.970686\n",
            "tmax_K    0.633246\n",
            "dtype: float64, tmin_K    0.466063\n",
            "tmax_K    0.361931\n",
            "dtype: float64, tmin_K    0.974769\n",
            "tmax_K    0.632569\n",
            "dtype: float64, tmin_K    0.978011\n",
            "tmax_K    0.627626\n",
            "dtype: float64, tmin_K    0.382851\n",
            "tmax_K    0.308338\n",
            "dtype: float64, tmin_K    0.979088\n",
            "tmax_K    0.634931\n",
            "dtype: float64, tmin_K    0.305968\n",
            "tmax_K    0.326398\n",
            "dtype: float64, tmin_K    0.987075\n",
            "tmax_K    0.635218\n",
            "dtype: float64, tmin_K    0.974502\n",
            "tmax_K    0.631047\n",
            "dtype: float64, tmin_K    0.976507\n",
            "tmax_K    0.633482\n",
            "dtype: float64, tmin_K    0.978060\n",
            "tmax_K    0.634132\n",
            "dtype: float64, tmin_K    0.977007\n",
            "tmax_K    0.633531\n",
            "dtype: float64, tmin_K    0.461137\n",
            "tmax_K    0.371716\n",
            "dtype: float64, tmin_K    0.454210\n",
            "tmax_K    0.367249\n",
            "dtype: float64, tmin_K    0.340547\n",
            "tmax_K    0.299952\n",
            "dtype: float64, tmin_K    0.395491\n",
            "tmax_K    0.302373\n",
            "dtype: float64, tmin_K    0.980705\n",
            "tmax_K    0.635496\n",
            "dtype: float64, tmin_K    0.338351\n",
            "tmax_K    0.359649\n",
            "dtype: float64, tmin_K    0.437244\n",
            "tmax_K    0.374736\n",
            "dtype: float64, tmin_K    0.988813\n",
            "tmax_K    0.631663\n",
            "dtype: float64, tmin_K    0.36297\n",
            "tmax_K    0.34031\n",
            "dtype: float64, tmin_K    0.357387\n",
            "tmax_K    0.264232\n",
            "dtype: float64, tmin_K    0.308316\n",
            "tmax_K    0.305389\n",
            "dtype: float64, tmin_K    0.981792\n",
            "tmax_K    0.626053\n",
            "dtype: float64, tmin_K    0.975541\n",
            "tmax_K    0.632568\n",
            "dtype: float64, tmin_K    0.968316\n",
            "tmax_K    0.634597\n",
            "dtype: float64, tmin_K    0.467982\n",
            "tmax_K    0.356662\n",
            "dtype: float64, tmin_K    0.975285\n",
            "tmax_K    0.630996\n",
            "dtype: float64, tmin_K    0.978419\n",
            "tmax_K    0.636193\n",
            "dtype: float64, tmin_K    0.981497\n",
            "tmax_K    0.636800\n",
            "dtype: float64, tmin_K    0.313749\n",
            "tmax_K    0.306200\n",
            "dtype: float64, tmin_K    0.982689\n",
            "tmax_K    0.638123\n",
            "dtype: float64, tmin_K    0.330252\n",
            "tmax_K    0.312692\n",
            "dtype: float64, tmin_K    0.981521\n",
            "tmax_K    0.635903\n",
            "dtype: float64, tmin_K    0.353547\n",
            "tmax_K    0.336769\n",
            "dtype: float64, tmin_K    0.976212\n",
            "tmax_K    0.632195\n",
            "dtype: float64, tmin_K    0.422608\n",
            "tmax_K    0.376753\n",
            "dtype: float64, tmin_K    0.973975\n",
            "tmax_K    0.633220\n",
            "dtype: float64, tmin_K    0.976044\n",
            "tmax_K    0.633949\n",
            "dtype: float64, tmin_K    0.363860\n",
            "tmax_K    0.327365\n",
            "dtype: float64, tmin_K    0.453949\n",
            "tmax_K    0.341606\n",
            "dtype: float64, tmin_K    0.449320\n",
            "tmax_K    0.359887\n",
            "dtype: float64, tmin_K    0.456126\n",
            "tmax_K    0.350853\n",
            "dtype: float64, tmin_K    0.455106\n",
            "tmax_K    0.324437\n",
            "dtype: float64, tmin_K    0.365982\n",
            "tmax_K    0.339302\n",
            "dtype: float64, tmin_K    0.438270\n",
            "tmax_K    0.362841\n",
            "dtype: float64, tmin_K    0.429728\n",
            "tmax_K    0.343519\n",
            "dtype: float64, tmin_K    0.353281\n",
            "tmax_K    0.313478\n",
            "dtype: float64, tmin_K    0.323391\n",
            "tmax_K    0.354663\n",
            "dtype: float64, tmin_K    0.982153\n",
            "tmax_K    0.636035\n",
            "dtype: float64, tmin_K    0.975825\n",
            "tmax_K    0.633433\n",
            "dtype: float64, tmin_K    0.425444\n",
            "tmax_K    0.319103\n",
            "dtype: float64, tmin_K    0.979445\n",
            "tmax_K    0.634201\n",
            "dtype: float64, tmin_K    0.971717\n",
            "tmax_K    0.631886\n",
            "dtype: float64, tmin_K    0.978361\n",
            "tmax_K    0.632329\n",
            "dtype: float64, tmin_K    0.976125\n",
            "tmax_K    0.639254\n",
            "dtype: float64, tmin_K    0.440360\n",
            "tmax_K    0.359743\n",
            "dtype: float64, tmin_K    0.368046\n",
            "tmax_K    0.305932\n",
            "dtype: float64, tmin_K    0.357575\n",
            "tmax_K    0.347460\n",
            "dtype: float64, tmin_K    0.388369\n",
            "tmax_K    0.326634\n",
            "dtype: float64, tmin_K    0.979147\n",
            "tmax_K    0.633396\n",
            "dtype: float64, tmin_K    0.978733\n",
            "tmax_K    0.634972\n",
            "dtype: float64, tmin_K    0.447676\n",
            "tmax_K    0.323205\n",
            "dtype: float64, tmin_K    0.289707\n",
            "tmax_K    0.378911\n",
            "dtype: float64, tmin_K    0.328734\n",
            "tmax_K    0.278747\n",
            "dtype: float64, tmin_K    0.975578\n",
            "tmax_K    0.630399\n",
            "dtype: float64, tmin_K    0.980854\n",
            "tmax_K    0.635246\n",
            "dtype: float64, tmin_K    0.974590\n",
            "tmax_K    0.631754\n",
            "dtype: float64, tmin_K    0.982663\n",
            "tmax_K    0.636759\n",
            "dtype: float64, tmin_K    0.333881\n",
            "tmax_K    0.357414\n",
            "dtype: float64, tmin_K    0.307684\n",
            "tmax_K    0.304865\n",
            "dtype: float64, tmin_K    0.338079\n",
            "tmax_K    0.363604\n",
            "dtype: float64, tmin_K    0.374191\n",
            "tmax_K    0.382758\n",
            "dtype: float64, tmin_K    0.326762\n",
            "tmax_K    0.333948\n",
            "dtype: float64, tmin_K    0.986447\n",
            "tmax_K    0.640421\n",
            "dtype: float64, tmin_K    0.977029\n",
            "tmax_K    0.633420\n",
            "dtype: float64, tmin_K    0.976859\n",
            "tmax_K    0.632719\n",
            "dtype: float64, tmin_K    0.341597\n",
            "tmax_K    0.285552\n",
            "dtype: float64, tmin_K    0.979394\n",
            "tmax_K    0.633072\n",
            "dtype: float64, tmin_K    0.979499\n",
            "tmax_K    0.633337\n",
            "dtype: float64, tmin_K    0.310325\n",
            "tmax_K    0.289266\n",
            "dtype: float64]\n",
            "\t BEST: 5 hidden layers, 50 hidden neurons\n",
            "Model Number: 29\n",
            "\t # Hidden Layers = 1\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.979093\n",
            "tmax_K    0.635177\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.974295\n",
            "tmax_K    0.630145\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.977768\n",
            "tmax_K    0.632797\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.980286\n",
            "tmax_K    0.635531\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.973415\n",
            "tmax_K    0.628534\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.976554\n",
            "tmax_K    0.632770\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.975255\n",
            "tmax_K    0.635010\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.960788\n",
            "tmax_K    0.630304\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.986061\n",
            "tmax_K    0.635702\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.981738\n",
            "tmax_K    0.636966\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.98519\n",
            "tmax_K    0.63892\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.980208\n",
            "tmax_K    0.635149\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.978518\n",
            "tmax_K    0.634526\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.978024\n",
            "tmax_K    0.628209\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.975860\n",
            "tmax_K    0.632682\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.976822\n",
            "tmax_K    0.629980\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.978095\n",
            "tmax_K    0.633733\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.986558\n",
            "tmax_K    0.636128\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.333049\n",
            "tmax_K    0.345790\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.979108\n",
            "tmax_K    0.633472\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.976331\n",
            "tmax_K    0.633007\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.971853\n",
            "tmax_K    0.632294\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.979279\n",
            "tmax_K    0.638751\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.972519\n",
            "tmax_K    0.633573\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.978668\n",
            "tmax_K    0.633718\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.980223\n",
            "tmax_K    0.635043\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.974132\n",
            "tmax_K    0.631944\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.975667\n",
            "tmax_K    0.629369\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.374916\n",
            "tmax_K    0.354394\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.355083\n",
            "tmax_K    0.353635\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.977016\n",
            "tmax_K    0.633835\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.976901\n",
            "tmax_K    0.632722\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.978958\n",
            "tmax_K    0.630665\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.984281\n",
            "tmax_K    0.636814\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.982859\n",
            "tmax_K    0.633875\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.974720\n",
            "tmax_K    0.631857\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.979938\n",
            "tmax_K    0.635597\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.987885\n",
            "tmax_K    0.636079\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.963201\n",
            "tmax_K    0.623648\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.981561\n",
            "tmax_K    0.637724\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.989174\n",
            "tmax_K    0.635644\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 2\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.411461\n",
            "tmax_K    0.336019\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.975395\n",
            "tmax_K    0.634292\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.976691\n",
            "tmax_K    0.631285\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.972683\n",
            "tmax_K    0.633689\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.350348\n",
            "tmax_K    0.357344\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.976164\n",
            "tmax_K    0.632898\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.982308\n",
            "tmax_K    0.637469\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.979511\n",
            "tmax_K    0.635234\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.972125\n",
            "tmax_K    0.636078\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.978829\n",
            "tmax_K    0.635485\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.979943\n",
            "tmax_K    0.636147\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.995667\n",
            "tmax_K    0.631293\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.970478\n",
            "tmax_K    0.633709\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.359767\n",
            "tmax_K    0.245406\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.966342\n",
            "tmax_K    0.628348\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.331728\n",
            "tmax_K    0.350232\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.981214\n",
            "tmax_K    0.633983\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.979171\n",
            "tmax_K    0.637268\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.978166\n",
            "tmax_K    0.634296\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.973989\n",
            "tmax_K    0.633010\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.482309\n",
            "tmax_K    0.354818\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.975786\n",
            "tmax_K    0.633184\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.985651\n",
            "tmax_K    0.635177\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.379554\n",
            "tmax_K    0.326507\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.965520\n",
            "tmax_K    0.628963\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.977675\n",
            "tmax_K    0.632870\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.975757\n",
            "tmax_K    0.635664\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.977734\n",
            "tmax_K    0.634121\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.975915\n",
            "tmax_K    0.630014\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.974194\n",
            "tmax_K    0.631229\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.977481\n",
            "tmax_K    0.633426\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.976761\n",
            "tmax_K    0.634107\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.305437\n",
            "tmax_K    0.360741\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.977325\n",
            "tmax_K    0.634165\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.977772\n",
            "tmax_K    0.634038\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.978797\n",
            "tmax_K    0.635997\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.973555\n",
            "tmax_K    0.628510\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.977436\n",
            "tmax_K    0.631470\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.984567\n",
            "tmax_K    0.634246\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.974732\n",
            "tmax_K    0.631881\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.977250\n",
            "tmax_K    0.634607\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 3\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.977757\n",
            "tmax_K    0.631276\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.970769\n",
            "tmax_K    0.632970\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.342070\n",
            "tmax_K    0.318853\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.977123\n",
            "tmax_K    0.633766\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.963965\n",
            "tmax_K    0.631478\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.980432\n",
            "tmax_K    0.635399\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.396967\n",
            "tmax_K    0.330446\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.427012\n",
            "tmax_K    0.316028\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.418948\n",
            "tmax_K    0.359635\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.978170\n",
            "tmax_K    0.634177\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.965070\n",
            "tmax_K    0.628519\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.980097\n",
            "tmax_K    0.634338\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.987624\n",
            "tmax_K    0.643766\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.976262\n",
            "tmax_K    0.636117\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.411051\n",
            "tmax_K    0.374499\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.980857\n",
            "tmax_K    0.634670\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.969616\n",
            "tmax_K    0.630392\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.975607\n",
            "tmax_K    0.632978\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.977671\n",
            "tmax_K    0.631620\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.446306\n",
            "tmax_K    0.351825\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.374683\n",
            "tmax_K    0.327219\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.331670\n",
            "tmax_K    0.295366\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.986516\n",
            "tmax_K    0.639114\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.459534\n",
            "tmax_K    0.368901\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.952726\n",
            "tmax_K    0.634326\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.979216\n",
            "tmax_K    0.634222\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.974388\n",
            "tmax_K    0.640976\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.344210\n",
            "tmax_K    0.348294\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.982537\n",
            "tmax_K    0.634969\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.379169\n",
            "tmax_K    0.305993\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.974272\n",
            "tmax_K    0.634818\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.978686\n",
            "tmax_K    0.632860\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.980864\n",
            "tmax_K    0.634467\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.978633\n",
            "tmax_K    0.635350\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.329111\n",
            "tmax_K    0.376756\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.988630\n",
            "tmax_K    0.639943\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.330027\n",
            "tmax_K    0.345998\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.365281\n",
            "tmax_K    0.333714\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.992268\n",
            "tmax_K    0.640144\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.979014\n",
            "tmax_K    0.633311\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.983777\n",
            "tmax_K    0.635703\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 4\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.463606\n",
            "tmax_K    0.349593\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.979502\n",
            "tmax_K    0.635666\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.404387\n",
            "tmax_K    0.326148\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.470965\n",
            "tmax_K    0.372730\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.975302\n",
            "tmax_K    0.632645\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.372104\n",
            "tmax_K    0.334665\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.344196\n",
            "tmax_K    0.280656\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.973839\n",
            "tmax_K    0.633647\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.459188\n",
            "tmax_K    0.362104\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.443139\n",
            "tmax_K    0.366570\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.329468\n",
            "tmax_K    0.314710\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.970482\n",
            "tmax_K    0.629917\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.974726\n",
            "tmax_K    0.639375\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.973982\n",
            "tmax_K    0.634641\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.983476\n",
            "tmax_K    0.632516\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.974291\n",
            "tmax_K    0.633096\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.447611\n",
            "tmax_K    0.363240\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.366649\n",
            "tmax_K    0.343679\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.415962\n",
            "tmax_K    0.310312\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.316415\n",
            "tmax_K    0.372072\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.973745\n",
            "tmax_K    0.632663\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.979971\n",
            "tmax_K    0.634738\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.982839\n",
            "tmax_K    0.630209\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.373436\n",
            "tmax_K    0.330076\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.314334\n",
            "tmax_K    0.311786\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.362888\n",
            "tmax_K    0.345513\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.979318\n",
            "tmax_K    0.634283\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.971822\n",
            "tmax_K    0.631763\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.979063\n",
            "tmax_K    0.635161\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.983201\n",
            "tmax_K    0.634322\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.394708\n",
            "tmax_K    0.316575\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.975415\n",
            "tmax_K    0.632619\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.329729\n",
            "tmax_K    0.334487\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.978830\n",
            "tmax_K    0.632507\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.329172\n",
            "tmax_K    0.338447\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.385250\n",
            "tmax_K    0.347879\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.977031\n",
            "tmax_K    0.634719\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.344832\n",
            "tmax_K    0.329467\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.979988\n",
            "tmax_K    0.634024\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.978404\n",
            "tmax_K    0.633406\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.961774\n",
            "tmax_K    0.634201\n",
            "dtype: float64\n",
            "\t # Hidden Layers = 5\n",
            "\t\t # hidden neurons = 10\n",
            "tmin_K    0.362863\n",
            "tmax_K    0.326275\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 11\n",
            "tmin_K    0.974492\n",
            "tmax_K    0.632540\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 12\n",
            "tmin_K    0.460012\n",
            "tmax_K    0.365350\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 13\n",
            "tmin_K    0.978527\n",
            "tmax_K    0.633129\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 14\n",
            "tmin_K    0.323610\n",
            "tmax_K    0.325057\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 15\n",
            "tmin_K    0.971837\n",
            "tmax_K    0.634288\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 16\n",
            "tmin_K    0.359121\n",
            "tmax_K    0.368902\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 17\n",
            "tmin_K    0.357482\n",
            "tmax_K    0.344812\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 18\n",
            "tmin_K    0.436401\n",
            "tmax_K    0.366113\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 19\n",
            "tmin_K    0.395060\n",
            "tmax_K    0.312562\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 20\n",
            "tmin_K    0.975599\n",
            "tmax_K    0.637577\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 21\n",
            "tmin_K    0.370000\n",
            "tmax_K    0.316971\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 22\n",
            "tmin_K    0.978321\n",
            "tmax_K    0.629483\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 23\n",
            "tmin_K    0.978274\n",
            "tmax_K    0.633840\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 24\n",
            "tmin_K    0.329465\n",
            "tmax_K    0.322396\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 25\n",
            "tmin_K    0.978473\n",
            "tmax_K    0.633819\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 26\n",
            "tmin_K    0.979709\n",
            "tmax_K    0.634120\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 27\n",
            "tmin_K    0.979208\n",
            "tmax_K    0.634033\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 28\n",
            "tmin_K    0.434931\n",
            "tmax_K    0.356628\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 29\n",
            "tmin_K    0.323267\n",
            "tmax_K    0.272761\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 30\n",
            "tmin_K    0.349083\n",
            "tmax_K    0.349292\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 31\n",
            "tmin_K    0.408204\n",
            "tmax_K    0.337391\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 32\n",
            "tmin_K    0.294831\n",
            "tmax_K    0.236991\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 33\n",
            "tmin_K    0.433734\n",
            "tmax_K    0.370654\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 34\n",
            "tmin_K    0.332163\n",
            "tmax_K    0.377634\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 35\n",
            "tmin_K    0.383739\n",
            "tmax_K    0.353732\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 36\n",
            "tmin_K    0.458477\n",
            "tmax_K    0.366153\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 37\n",
            "tmin_K    0.370920\n",
            "tmax_K    0.358692\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 38\n",
            "tmin_K    0.353358\n",
            "tmax_K    0.376873\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 39\n",
            "tmin_K    0.359162\n",
            "tmax_K    0.326476\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 40\n",
            "tmin_K    0.976618\n",
            "tmax_K    0.633003\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 41\n",
            "tmin_K    0.976494\n",
            "tmax_K    0.633103\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 42\n",
            "tmin_K    0.365022\n",
            "tmax_K    0.376885\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 43\n",
            "tmin_K    0.978331\n",
            "tmax_K    0.633962\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 44\n",
            "tmin_K    0.291253\n",
            "tmax_K    0.296629\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 45\n",
            "tmin_K    0.321834\n",
            "tmax_K    0.364056\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 46\n",
            "tmin_K    0.965387\n",
            "tmax_K    0.627764\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 47\n",
            "tmin_K    0.435537\n",
            "tmax_K    0.325447\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 48\n",
            "tmin_K    0.294515\n",
            "tmax_K    0.279365\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 49\n",
            "tmin_K    0.487067\n",
            "tmax_K    0.354586\n",
            "dtype: float64\n",
            "\t\t # hidden neurons = 50\n",
            "tmin_K    0.977253\n",
            "tmax_K    0.633907\n",
            "dtype: float64\n",
            "[1.61427008 1.60444046 1.61056521 1.61581684 1.60194872 1.60932428\n",
            " 1.61026479 1.59109224 1.62176296 1.61870305 1.62411012 1.61535773\n",
            " 1.61304444 1.60623239 1.60854212 1.60680225 1.61182798 1.62268558\n",
            " 0.6788391  1.61257991 1.60933737 1.60414747 1.61802978 1.60609179\n",
            " 1.61238634 1.61526675 1.60607516 1.60503674 0.72930987 0.70871829\n",
            " 1.61085095 1.60962293 1.6096226  1.62109505 1.61673447 1.60657733\n",
            " 1.61553544 1.62396345 1.5868494  1.61928477 1.62481859 0.74747984\n",
            " 1.6096871  1.60797641 1.60637167 0.70769137 1.60906264 1.6197763\n",
            " 1.61474568 1.60820222 1.61431449 1.61609035 1.62695973 1.60418692\n",
            " 0.60517326 1.59469005 0.68195999 1.61519702 1.61643839 1.61246146\n",
            " 1.60699921 0.83712769 1.60896954 1.62082742 0.70606108 1.59448299\n",
            " 1.61054476 1.61142059 1.61185503 1.60592862 1.60542351 1.61090724\n",
            " 1.61086823 0.66617838 1.61148942 1.61180979 1.61479353 1.60206428\n",
            " 1.6089058  1.61881228 1.60661292 1.61185723 1.60903301 1.60373909\n",
            " 0.66092297 1.6108884  1.59544299 1.61583065 0.72741373 0.74304079\n",
            " 0.77858353 1.61234721 1.59358872 1.61443435 1.63139006 1.61237969\n",
            " 0.78554939 1.61552718 1.60000803 1.60858525 1.60929122 0.79813119\n",
            " 0.70190202 0.62703639 1.62562977 0.82843493 1.5870519  1.61343795\n",
            " 1.61536397 0.69250409 1.61750588 0.68516181 1.60909008 1.61154663\n",
            " 1.61533168 1.61398366 0.70586612 1.628573   0.67602479 0.69899544\n",
            " 1.63241195 1.61232547 1.61947921 0.8131984  1.61516812 0.73053574\n",
            " 0.84369475 1.60794653 0.7067688  0.62485232 1.60748641 0.82129173\n",
            " 0.80970957 0.64417846 1.60039898 1.61410067 1.60862296 1.61599215\n",
            " 1.60738739 0.81085153 0.71032846 0.72627382 0.68848647 1.60640763\n",
            " 1.6147095  1.61304762 0.70351226 0.6261191  0.70840082 1.61360181\n",
            " 1.60358512 1.61422387 1.61752313 0.71128315 1.60803354 0.66421671\n",
            " 1.61133708 0.66761842 0.73312831 1.61174949 0.67429936 1.61401189\n",
            " 1.61180984 1.59597501 0.68913796 1.6070311  0.82536173 1.61165549\n",
            " 0.64866755 1.60612471 0.72802283 0.70229445 0.80251395 0.70762254\n",
            " 1.61317562 0.68697157 1.60780439 1.61211341 0.65186172 1.61229183\n",
            " 1.61382904 1.61324161 0.79155891 0.59602829 0.69837569 0.74559532\n",
            " 0.53182191 0.80438822 0.70979744 0.73747035 0.82463034 0.72961154\n",
            " 0.73023097 0.68563802 1.60962098 1.60959746 0.74190677 1.61229365\n",
            " 0.58788162 0.68588964 1.59315126 0.76098324 0.57387955 0.84165302\n",
            " 1.61115928]\n",
            "[tmin_K    0.979093\n",
            "tmax_K    0.635177\n",
            "dtype: float64, tmin_K    0.974295\n",
            "tmax_K    0.630145\n",
            "dtype: float64, tmin_K    0.977768\n",
            "tmax_K    0.632797\n",
            "dtype: float64, tmin_K    0.980286\n",
            "tmax_K    0.635531\n",
            "dtype: float64, tmin_K    0.973415\n",
            "tmax_K    0.628534\n",
            "dtype: float64, tmin_K    0.976554\n",
            "tmax_K    0.632770\n",
            "dtype: float64, tmin_K    0.975255\n",
            "tmax_K    0.635010\n",
            "dtype: float64, tmin_K    0.960788\n",
            "tmax_K    0.630304\n",
            "dtype: float64, tmin_K    0.986061\n",
            "tmax_K    0.635702\n",
            "dtype: float64, tmin_K    0.981738\n",
            "tmax_K    0.636966\n",
            "dtype: float64, tmin_K    0.98519\n",
            "tmax_K    0.63892\n",
            "dtype: float64, tmin_K    0.980208\n",
            "tmax_K    0.635149\n",
            "dtype: float64, tmin_K    0.978518\n",
            "tmax_K    0.634526\n",
            "dtype: float64, tmin_K    0.978024\n",
            "tmax_K    0.628209\n",
            "dtype: float64, tmin_K    0.975860\n",
            "tmax_K    0.632682\n",
            "dtype: float64, tmin_K    0.976822\n",
            "tmax_K    0.629980\n",
            "dtype: float64, tmin_K    0.978095\n",
            "tmax_K    0.633733\n",
            "dtype: float64, tmin_K    0.986558\n",
            "tmax_K    0.636128\n",
            "dtype: float64, tmin_K    0.333049\n",
            "tmax_K    0.345790\n",
            "dtype: float64, tmin_K    0.979108\n",
            "tmax_K    0.633472\n",
            "dtype: float64, tmin_K    0.976331\n",
            "tmax_K    0.633007\n",
            "dtype: float64, tmin_K    0.971853\n",
            "tmax_K    0.632294\n",
            "dtype: float64, tmin_K    0.979279\n",
            "tmax_K    0.638751\n",
            "dtype: float64, tmin_K    0.972519\n",
            "tmax_K    0.633573\n",
            "dtype: float64, tmin_K    0.978668\n",
            "tmax_K    0.633718\n",
            "dtype: float64, tmin_K    0.980223\n",
            "tmax_K    0.635043\n",
            "dtype: float64, tmin_K    0.974132\n",
            "tmax_K    0.631944\n",
            "dtype: float64, tmin_K    0.975667\n",
            "tmax_K    0.629369\n",
            "dtype: float64, tmin_K    0.374916\n",
            "tmax_K    0.354394\n",
            "dtype: float64, tmin_K    0.355083\n",
            "tmax_K    0.353635\n",
            "dtype: float64, tmin_K    0.977016\n",
            "tmax_K    0.633835\n",
            "dtype: float64, tmin_K    0.976901\n",
            "tmax_K    0.632722\n",
            "dtype: float64, tmin_K    0.978958\n",
            "tmax_K    0.630665\n",
            "dtype: float64, tmin_K    0.984281\n",
            "tmax_K    0.636814\n",
            "dtype: float64, tmin_K    0.982859\n",
            "tmax_K    0.633875\n",
            "dtype: float64, tmin_K    0.974720\n",
            "tmax_K    0.631857\n",
            "dtype: float64, tmin_K    0.979938\n",
            "tmax_K    0.635597\n",
            "dtype: float64, tmin_K    0.987885\n",
            "tmax_K    0.636079\n",
            "dtype: float64, tmin_K    0.963201\n",
            "tmax_K    0.623648\n",
            "dtype: float64, tmin_K    0.981561\n",
            "tmax_K    0.637724\n",
            "dtype: float64, tmin_K    0.989174\n",
            "tmax_K    0.635644\n",
            "dtype: float64, tmin_K    0.411461\n",
            "tmax_K    0.336019\n",
            "dtype: float64, tmin_K    0.975395\n",
            "tmax_K    0.634292\n",
            "dtype: float64, tmin_K    0.976691\n",
            "tmax_K    0.631285\n",
            "dtype: float64, tmin_K    0.972683\n",
            "tmax_K    0.633689\n",
            "dtype: float64, tmin_K    0.350348\n",
            "tmax_K    0.357344\n",
            "dtype: float64, tmin_K    0.976164\n",
            "tmax_K    0.632898\n",
            "dtype: float64, tmin_K    0.982308\n",
            "tmax_K    0.637469\n",
            "dtype: float64, tmin_K    0.979511\n",
            "tmax_K    0.635234\n",
            "dtype: float64, tmin_K    0.972125\n",
            "tmax_K    0.636078\n",
            "dtype: float64, tmin_K    0.978829\n",
            "tmax_K    0.635485\n",
            "dtype: float64, tmin_K    0.979943\n",
            "tmax_K    0.636147\n",
            "dtype: float64, tmin_K    0.995667\n",
            "tmax_K    0.631293\n",
            "dtype: float64, tmin_K    0.970478\n",
            "tmax_K    0.633709\n",
            "dtype: float64, tmin_K    0.359767\n",
            "tmax_K    0.245406\n",
            "dtype: float64, tmin_K    0.966342\n",
            "tmax_K    0.628348\n",
            "dtype: float64, tmin_K    0.331728\n",
            "tmax_K    0.350232\n",
            "dtype: float64, tmin_K    0.981214\n",
            "tmax_K    0.633983\n",
            "dtype: float64, tmin_K    0.979171\n",
            "tmax_K    0.637268\n",
            "dtype: float64, tmin_K    0.978166\n",
            "tmax_K    0.634296\n",
            "dtype: float64, tmin_K    0.973989\n",
            "tmax_K    0.633010\n",
            "dtype: float64, tmin_K    0.482309\n",
            "tmax_K    0.354818\n",
            "dtype: float64, tmin_K    0.975786\n",
            "tmax_K    0.633184\n",
            "dtype: float64, tmin_K    0.985651\n",
            "tmax_K    0.635177\n",
            "dtype: float64, tmin_K    0.379554\n",
            "tmax_K    0.326507\n",
            "dtype: float64, tmin_K    0.965520\n",
            "tmax_K    0.628963\n",
            "dtype: float64, tmin_K    0.977675\n",
            "tmax_K    0.632870\n",
            "dtype: float64, tmin_K    0.975757\n",
            "tmax_K    0.635664\n",
            "dtype: float64, tmin_K    0.977734\n",
            "tmax_K    0.634121\n",
            "dtype: float64, tmin_K    0.975915\n",
            "tmax_K    0.630014\n",
            "dtype: float64, tmin_K    0.974194\n",
            "tmax_K    0.631229\n",
            "dtype: float64, tmin_K    0.977481\n",
            "tmax_K    0.633426\n",
            "dtype: float64, tmin_K    0.976761\n",
            "tmax_K    0.634107\n",
            "dtype: float64, tmin_K    0.305437\n",
            "tmax_K    0.360741\n",
            "dtype: float64, tmin_K    0.977325\n",
            "tmax_K    0.634165\n",
            "dtype: float64, tmin_K    0.977772\n",
            "tmax_K    0.634038\n",
            "dtype: float64, tmin_K    0.978797\n",
            "tmax_K    0.635997\n",
            "dtype: float64, tmin_K    0.973555\n",
            "tmax_K    0.628510\n",
            "dtype: float64, tmin_K    0.977436\n",
            "tmax_K    0.631470\n",
            "dtype: float64, tmin_K    0.984567\n",
            "tmax_K    0.634246\n",
            "dtype: float64, tmin_K    0.974732\n",
            "tmax_K    0.631881\n",
            "dtype: float64, tmin_K    0.977250\n",
            "tmax_K    0.634607\n",
            "dtype: float64, tmin_K    0.977757\n",
            "tmax_K    0.631276\n",
            "dtype: float64, tmin_K    0.970769\n",
            "tmax_K    0.632970\n",
            "dtype: float64, tmin_K    0.342070\n",
            "tmax_K    0.318853\n",
            "dtype: float64, tmin_K    0.977123\n",
            "tmax_K    0.633766\n",
            "dtype: float64, tmin_K    0.963965\n",
            "tmax_K    0.631478\n",
            "dtype: float64, tmin_K    0.980432\n",
            "tmax_K    0.635399\n",
            "dtype: float64, tmin_K    0.396967\n",
            "tmax_K    0.330446\n",
            "dtype: float64, tmin_K    0.427012\n",
            "tmax_K    0.316028\n",
            "dtype: float64, tmin_K    0.418948\n",
            "tmax_K    0.359635\n",
            "dtype: float64, tmin_K    0.978170\n",
            "tmax_K    0.634177\n",
            "dtype: float64, tmin_K    0.965070\n",
            "tmax_K    0.628519\n",
            "dtype: float64, tmin_K    0.980097\n",
            "tmax_K    0.634338\n",
            "dtype: float64, tmin_K    0.987624\n",
            "tmax_K    0.643766\n",
            "dtype: float64, tmin_K    0.976262\n",
            "tmax_K    0.636117\n",
            "dtype: float64, tmin_K    0.411051\n",
            "tmax_K    0.374499\n",
            "dtype: float64, tmin_K    0.980857\n",
            "tmax_K    0.634670\n",
            "dtype: float64, tmin_K    0.969616\n",
            "tmax_K    0.630392\n",
            "dtype: float64, tmin_K    0.975607\n",
            "tmax_K    0.632978\n",
            "dtype: float64, tmin_K    0.977671\n",
            "tmax_K    0.631620\n",
            "dtype: float64, tmin_K    0.446306\n",
            "tmax_K    0.351825\n",
            "dtype: float64, tmin_K    0.374683\n",
            "tmax_K    0.327219\n",
            "dtype: float64, tmin_K    0.331670\n",
            "tmax_K    0.295366\n",
            "dtype: float64, tmin_K    0.986516\n",
            "tmax_K    0.639114\n",
            "dtype: float64, tmin_K    0.459534\n",
            "tmax_K    0.368901\n",
            "dtype: float64, tmin_K    0.952726\n",
            "tmax_K    0.634326\n",
            "dtype: float64, tmin_K    0.979216\n",
            "tmax_K    0.634222\n",
            "dtype: float64, tmin_K    0.974388\n",
            "tmax_K    0.640976\n",
            "dtype: float64, tmin_K    0.344210\n",
            "tmax_K    0.348294\n",
            "dtype: float64, tmin_K    0.982537\n",
            "tmax_K    0.634969\n",
            "dtype: float64, tmin_K    0.379169\n",
            "tmax_K    0.305993\n",
            "dtype: float64, tmin_K    0.974272\n",
            "tmax_K    0.634818\n",
            "dtype: float64, tmin_K    0.978686\n",
            "tmax_K    0.632860\n",
            "dtype: float64, tmin_K    0.980864\n",
            "tmax_K    0.634467\n",
            "dtype: float64, tmin_K    0.978633\n",
            "tmax_K    0.635350\n",
            "dtype: float64, tmin_K    0.329111\n",
            "tmax_K    0.376756\n",
            "dtype: float64, tmin_K    0.988630\n",
            "tmax_K    0.639943\n",
            "dtype: float64, tmin_K    0.330027\n",
            "tmax_K    0.345998\n",
            "dtype: float64, tmin_K    0.365281\n",
            "tmax_K    0.333714\n",
            "dtype: float64, tmin_K    0.992268\n",
            "tmax_K    0.640144\n",
            "dtype: float64, tmin_K    0.979014\n",
            "tmax_K    0.633311\n",
            "dtype: float64, tmin_K    0.983777\n",
            "tmax_K    0.635703\n",
            "dtype: float64, tmin_K    0.463606\n",
            "tmax_K    0.349593\n",
            "dtype: float64, tmin_K    0.979502\n",
            "tmax_K    0.635666\n",
            "dtype: float64, tmin_K    0.404387\n",
            "tmax_K    0.326148\n",
            "dtype: float64, tmin_K    0.470965\n",
            "tmax_K    0.372730\n",
            "dtype: float64, tmin_K    0.975302\n",
            "tmax_K    0.632645\n",
            "dtype: float64, tmin_K    0.372104\n",
            "tmax_K    0.334665\n",
            "dtype: float64, tmin_K    0.344196\n",
            "tmax_K    0.280656\n",
            "dtype: float64, tmin_K    0.973839\n",
            "tmax_K    0.633647\n",
            "dtype: float64, tmin_K    0.459188\n",
            "tmax_K    0.362104\n",
            "dtype: float64, tmin_K    0.443139\n",
            "tmax_K    0.366570\n",
            "dtype: float64, tmin_K    0.329468\n",
            "tmax_K    0.314710\n",
            "dtype: float64, tmin_K    0.970482\n",
            "tmax_K    0.629917\n",
            "dtype: float64, tmin_K    0.974726\n",
            "tmax_K    0.639375\n",
            "dtype: float64, tmin_K    0.973982\n",
            "tmax_K    0.634641\n",
            "dtype: float64, tmin_K    0.983476\n",
            "tmax_K    0.632516\n",
            "dtype: float64, tmin_K    0.974291\n",
            "tmax_K    0.633096\n",
            "dtype: float64, tmin_K    0.447611\n",
            "tmax_K    0.363240\n",
            "dtype: float64, tmin_K    0.366649\n",
            "tmax_K    0.343679\n",
            "dtype: float64, tmin_K    0.415962\n",
            "tmax_K    0.310312\n",
            "dtype: float64, tmin_K    0.316415\n",
            "tmax_K    0.372072\n",
            "dtype: float64, tmin_K    0.973745\n",
            "tmax_K    0.632663\n",
            "dtype: float64, tmin_K    0.979971\n",
            "tmax_K    0.634738\n",
            "dtype: float64, tmin_K    0.982839\n",
            "tmax_K    0.630209\n",
            "dtype: float64, tmin_K    0.373436\n",
            "tmax_K    0.330076\n",
            "dtype: float64, tmin_K    0.314334\n",
            "tmax_K    0.311786\n",
            "dtype: float64, tmin_K    0.362888\n",
            "tmax_K    0.345513\n",
            "dtype: float64, tmin_K    0.979318\n",
            "tmax_K    0.634283\n",
            "dtype: float64, tmin_K    0.971822\n",
            "tmax_K    0.631763\n",
            "dtype: float64, tmin_K    0.979063\n",
            "tmax_K    0.635161\n",
            "dtype: float64, tmin_K    0.983201\n",
            "tmax_K    0.634322\n",
            "dtype: float64, tmin_K    0.394708\n",
            "tmax_K    0.316575\n",
            "dtype: float64, tmin_K    0.975415\n",
            "tmax_K    0.632619\n",
            "dtype: float64, tmin_K    0.329729\n",
            "tmax_K    0.334487\n",
            "dtype: float64, tmin_K    0.978830\n",
            "tmax_K    0.632507\n",
            "dtype: float64, tmin_K    0.329172\n",
            "tmax_K    0.338447\n",
            "dtype: float64, tmin_K    0.385250\n",
            "tmax_K    0.347879\n",
            "dtype: float64, tmin_K    0.977031\n",
            "tmax_K    0.634719\n",
            "dtype: float64, tmin_K    0.344832\n",
            "tmax_K    0.329467\n",
            "dtype: float64, tmin_K    0.979988\n",
            "tmax_K    0.634024\n",
            "dtype: float64, tmin_K    0.978404\n",
            "tmax_K    0.633406\n",
            "dtype: float64, tmin_K    0.961774\n",
            "tmax_K    0.634201\n",
            "dtype: float64, tmin_K    0.362863\n",
            "tmax_K    0.326275\n",
            "dtype: float64, tmin_K    0.974492\n",
            "tmax_K    0.632540\n",
            "dtype: float64, tmin_K    0.460012\n",
            "tmax_K    0.365350\n",
            "dtype: float64, tmin_K    0.978527\n",
            "tmax_K    0.633129\n",
            "dtype: float64, tmin_K    0.323610\n",
            "tmax_K    0.325057\n",
            "dtype: float64, tmin_K    0.971837\n",
            "tmax_K    0.634288\n",
            "dtype: float64, tmin_K    0.359121\n",
            "tmax_K    0.368902\n",
            "dtype: float64, tmin_K    0.357482\n",
            "tmax_K    0.344812\n",
            "dtype: float64, tmin_K    0.436401\n",
            "tmax_K    0.366113\n",
            "dtype: float64, tmin_K    0.395060\n",
            "tmax_K    0.312562\n",
            "dtype: float64, tmin_K    0.975599\n",
            "tmax_K    0.637577\n",
            "dtype: float64, tmin_K    0.370000\n",
            "tmax_K    0.316971\n",
            "dtype: float64, tmin_K    0.978321\n",
            "tmax_K    0.629483\n",
            "dtype: float64, tmin_K    0.978274\n",
            "tmax_K    0.633840\n",
            "dtype: float64, tmin_K    0.329465\n",
            "tmax_K    0.322396\n",
            "dtype: float64, tmin_K    0.978473\n",
            "tmax_K    0.633819\n",
            "dtype: float64, tmin_K    0.979709\n",
            "tmax_K    0.634120\n",
            "dtype: float64, tmin_K    0.979208\n",
            "tmax_K    0.634033\n",
            "dtype: float64, tmin_K    0.434931\n",
            "tmax_K    0.356628\n",
            "dtype: float64, tmin_K    0.323267\n",
            "tmax_K    0.272761\n",
            "dtype: float64, tmin_K    0.349083\n",
            "tmax_K    0.349292\n",
            "dtype: float64, tmin_K    0.408204\n",
            "tmax_K    0.337391\n",
            "dtype: float64, tmin_K    0.294831\n",
            "tmax_K    0.236991\n",
            "dtype: float64, tmin_K    0.433734\n",
            "tmax_K    0.370654\n",
            "dtype: float64, tmin_K    0.332163\n",
            "tmax_K    0.377634\n",
            "dtype: float64, tmin_K    0.383739\n",
            "tmax_K    0.353732\n",
            "dtype: float64, tmin_K    0.458477\n",
            "tmax_K    0.366153\n",
            "dtype: float64, tmin_K    0.370920\n",
            "tmax_K    0.358692\n",
            "dtype: float64, tmin_K    0.353358\n",
            "tmax_K    0.376873\n",
            "dtype: float64, tmin_K    0.359162\n",
            "tmax_K    0.326476\n",
            "dtype: float64, tmin_K    0.976618\n",
            "tmax_K    0.633003\n",
            "dtype: float64, tmin_K    0.976494\n",
            "tmax_K    0.633103\n",
            "dtype: float64, tmin_K    0.365022\n",
            "tmax_K    0.376885\n",
            "dtype: float64, tmin_K    0.978331\n",
            "tmax_K    0.633962\n",
            "dtype: float64, tmin_K    0.291253\n",
            "tmax_K    0.296629\n",
            "dtype: float64, tmin_K    0.321834\n",
            "tmax_K    0.364056\n",
            "dtype: float64, tmin_K    0.965387\n",
            "tmax_K    0.627764\n",
            "dtype: float64, tmin_K    0.435537\n",
            "tmax_K    0.325447\n",
            "dtype: float64, tmin_K    0.294515\n",
            "tmax_K    0.279365\n",
            "dtype: float64, tmin_K    0.487067\n",
            "tmax_K    0.354586\n",
            "dtype: float64, tmin_K    0.977253\n",
            "tmax_K    0.633907\n",
            "dtype: float64]\n",
            "\t BEST: 5 hidden layers, 32 hidden neurons\n",
            "\tEnsemble: 26e21445-845e-49b0-942a-5a31bbddcad0 \n",
            "\tRMSE:tmin_K    0.287897\n",
            "tmax_K    0.258562\n",
            "dtype: float64\n"
          ]
        }
      ],
      "source": [
        "ensemble_guid = uuid.uuid4()\n",
        "max_iter = 1500 #max number of epochs to run\n",
        "early_stopping = True #True = stop early if validation error begins to rise\n",
        "\n",
        "y_out_all_nhn = []\n",
        "y_out_ensemble = []\n",
        "RMSE_ensemble = [] #RMSE for each model in the ensemble\n",
        "RMSE_ensemble_cumsum = [] #RMSE of the cumulative saltation for each model\n",
        "nhn_best = []\n",
        "nhl_best = []\n",
        "all_models = []    ### NEW\n",
        "\n",
        "for model_num in range(hyperparams['num_models']): #for each model in the ensemble\n",
        "    print('Model Number: ' + str(model_num))\n",
        "    RMSE = []\n",
        "    y_out_all_nhn = []\n",
        "    nhn = []\n",
        "    nhl = []\n",
        "    models = []   ### NEW\n",
        "\n",
        "    for num_hidden_layers in range(1,hyperparams['max_hidden_layers']+1):\n",
        "        print('\\t # Hidden Layers = ' + str(num_hidden_layers))\n",
        "        for num_hidden_neurons in range(hyperparams['min_nhn'],hyperparams['max_nhn']+1): #for each number of hidden neurons\n",
        "            print('\\t\\t # hidden neurons = ' + str(num_hidden_neurons))\n",
        "\n",
        "            hidden_layer_sizes = (num_hidden_neurons,num_hidden_layers)\n",
        "            model = MLPRegressor(hidden_layer_sizes=hidden_layer_sizes,\n",
        "                                 verbose=False,\n",
        "                                 max_iter=max_iter,\n",
        "                                 early_stopping = early_stopping,\n",
        "                                 validation_fraction = hyperparams['validation_fraction'],\n",
        "                                 batch_size =hyperparams['batch_size'],\n",
        "                                 solver = hyperparams['solver'],\n",
        "                                 alpha= hyperparams['alpha'], # regularize in order to penalize outliers\n",
        "                                 # so far .01 reg has lowest RMSE\n",
        "                                 activation = hyperparams['activation'],\n",
        "                                 learning_rate_init = hyperparams['learning_rate_init'],\n",
        "                                 learning_rate = hyperparams['learning_rate'],\n",
        "                                 shuffle = hyperparams['shuffling'])\n",
        "\n",
        "            model.fit(x_train,y_train) #train the model\n",
        "\n",
        "            y_out_this_nhn = model.predict(x_val) #model prediction for this number of hidden neurons (nhn)\n",
        "            y_out_all_nhn.append(y_out_this_nhn) #store all models -- will select best one best on RMSE\n",
        "\n",
        "            RMSE.append(rmse(y_val,y_out_this_nhn)) #RMSE between cumulative curves\n",
        "            print(rmse(y_val,y_out_this_nhn))\n",
        "            nhn.append(num_hidden_neurons)\n",
        "            nhl.append(num_hidden_layers)\n",
        "\n",
        "            models.append(model) ### NEW\n",
        "\n",
        "    #index of model with lowest RMSE\n",
        "    tmin_tmax_total_rmses = np.sum(RMSE, axis=1)\n",
        "    print(tmin_tmax_total_rmses)\n",
        "    print(RMSE)\n",
        "    indBest = tmin_tmax_total_rmses.argmin()#((np.min(tmin_tmax_total_rmses)))\n",
        "    RMSE_ensemble.append(np.min(tmin_tmax_total_rmses))\n",
        "    nhn_best.append(nhn[indBest])\n",
        "    nhl_best.append(nhl[indBest])\n",
        "\n",
        "    all_models.append(models[indBest])\n",
        "    y_out_ensemble.append(y_out_all_nhn[indBest])\n",
        "    print('\\t BEST: ' + str(nhl_best[model_num]) + ' hidden layers, '+ str(nhn_best[model_num]) + ' hidden neurons')\n",
        "\n",
        "\n",
        "y_out_ensemble_mean = np.mean(y_out_ensemble,axis=0)\n",
        "RMSE_ensemble_mean = rmse(y_out_ensemble_mean, y_val)\n",
        "print(f\"\\tEnsemble: {ensemble_guid} \\r\\n\\t\" +\n",
        "      f\"RMSE:{str(RMSE_ensemble_mean)}\")\n",
        "hyperparams['rmse'] = RMSE_ensemble_mean\n",
        "hyperparams['ensemble_guid'] = ensemble_guid\n",
        "\n",
        "\n",
        "wdir = '/content/drive/MyDrive/Colab Notebooks/EOSC510_Final'\n",
        "\n",
        "### Now, we have a list of the best-performing models.\n",
        "### First thing to do is save them.\n",
        "### This is really important if you're working with a partner & in the interest of making results repeatable\n",
        "### and just so you don't have to run the above code all the time\n",
        "pickle_path = f\"{wdir}/{ensemble_guid}.p\"\n",
        "pickle.dump(all_models,open(pickle_path,\"wb\"))\n",
        "\n",
        "# Save results to a CSV to track which architecture resulted in lowest RMSE\n",
        "with open(f\"{wdir}/final_mlp_hyperparams_log.csv\", 'a', newline='') as file:\n",
        "    writer = csv.DictWriter(file, fieldnames=hyperparams.keys())\n",
        "    if file.tell() == 0:  # Write header if file is empty\n",
        "        writer.writeheader()\n",
        "    writer.writerow(hyperparams)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## vis models"
      ],
      "metadata": {
        "id": "XnpszIs_YauU"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(12,8))\n",
        "\n",
        "plt.subplot(221)\n",
        "plt.scatter(range(len(nhn_best)),nhn_best)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('# Hidden Neurons')\n",
        "plt.title('Hidden Neurons')\n",
        "plt.subplot(222)\n",
        "plt.scatter(range(len(nhl_best)),nhl_best)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('# Hidden Layers')\n",
        "plt.title('Hidden Layers')\n",
        "plt.tight_layout()\n",
        "\n",
        "\n",
        "# plt.figure(figsize=(12,8))\n",
        "plt.scatter(len(RMSE_ensemble),RMSE_ensemble_mean,c='k',marker='*')\n",
        "plt.scatter(range(len(RMSE_ensemble)),RMSE_ensemble)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('RMSE')\n",
        "plt.title('Error')\n",
        "\n",
        "# plt.subplot(244)\n",
        "# plt.scatter(y_val,y_out_ensemble_mean)\n",
        "# plt.xlabel('y_test')\n",
        "# plt.ylabel('predicted y')\n",
        "# #plt.plot((np.min(y_test),np.max(y_test)),'k--')\n",
        "# # plt.xlabel('y_test')\n",
        "# # plt.ylabel('y_model')\n",
        "# # plt.title('Ensemble')\n",
        "# # plt.subplot(212)\n",
        "# # plt.plot(y_out_ensemble_mean, label='ensemble mean')\n",
        "# # plt.plot(np.array(y_val),alpha = 0.5, label='y_true')\n",
        "plt.tight_layout()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 493
        },
        "id": "_UH4Zcy6YubU",
        "outputId": "3281d1fb-3df6-4a0c-ef88-17d54502b631"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "AttributeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-12-3ac5ad732262>\u001b[0m in \u001b[0;36m<cell line: 17>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     15\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;31m# plt.figure(figsize=(12,8))\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mRMSE_ensemble_mean\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'k'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmarker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'*'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     18\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mxlabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Model #'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mAttributeError\u001b[0m: 'list' object has no attribute 'mean'"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABKQAAAGuCAYAAACugN4RAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAABwqElEQVR4nO3dfVxUZf7/8feACt7AKBTOmKikphKZYWGkqXmTaJGu7rfyJm8yK1crdbtjt0K6Q2t307aiu013M7O1rVy7wSwVt9QwkZQoN11MK5BWckANNDi/P/gx6wgoA8OcYXg9H495PJzrXHPOZ645Ux8+c53rWAzDMAQAAAAAAAB4SYDZAQAAAAAAAKB5oSAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAWgTrp166bp06eftd/y5ctlsVi0f/9+j+0TAADAn5FnAWiOKEgBzVBVMvP555/XuH3o0KGKiYnxclTes2nTJlksFlksFu3YsaPa9unTp6tdu3YmRAYAAJo68qzKPOvNN980OxQAPq6F2QEAaBr27NmjgAD/q2EvXLhQa9euNTsMAADQjPlrngUAZ8J/9QDUSVBQkFq2bGl2GB7Vr18/vfvuu8rKyjI7FEnSL7/8ohMnTpgdBgAA8DJ/zLOaCsMw9PPPP5sdBtAsUZACUCc1rUPw5ZdfatiwYWrdurU6d+6sRx99VBUVFdVeaxiGHn30UXXu3Flt2rTRVVddpS+//LLG4xw5ckTz5s1TZGSkgoKC1KNHDy1evNhlv/v375fFYtEf/vAHvfjii+revbuCgoJ02WWXafv27XV+T3fccYc6dOighQsX1qn/Bx98oCuvvFJt27ZVSEiIrrnmmmrvY+jQoRo6dGi1106fPl3dunWr8T0sWbLE+R5yc3MlSRs2bHAeq3379ho7dqy++uorl30uXLhQFotFe/fu1fTp09W+fXtZrVbNmDFDx48fd+m7fv16DRo0SO3bt1e7du3Uq1cv/e53v6vT+wYAAI3LH/Oss/nDH/6gK664QuHh4WrdurX69+9f7TK/IUOG6OKLL67x9b169dKoUaOczysqKrRkyRJdeOGFCg4OVseOHXXbbbfpp59+cnldt27ddO2112rdunW69NJL1bp1a73wwguSyJcAb+OSPaAZczgc+u9//1ut/eTJk2d9bUFBga666ir98ssvuv/++9W2bVu9+OKLat26dbW+Dz30kB599FGNGTNGY8aMUVZWlq6++upqs4GOHz+uIUOG6Pvvv9dtt92mLl26aMuWLUpKSlJ+fr6WLFni0n/lypUqKSnRbbfdJovFoieeeELjx4/Xf/7znzr9yhgaGqr58+froYceUlZWlmJjY2vt++qrr2ratGkaNWqUFi9erOPHjystLU2DBg3Szp07XYpN7li2bJlKS0t16623KigoSGFhYfroo480evRonX/++Vq4cKF+/vln/fnPf9bAgQOVlZVV7VjXX3+9oqKilJqaqqysLL388suKiIjQ4sWLJVUmtNdee6369u2rhx9+WEFBQdq7d68+/fTTesUMAADOrrnnWWezdOlSXXfddZo8ebJOnDihVatW6f/+7//07rvv6pprrpEk3XTTTZo1a5ZycnJc1t3avn27/v3vf+uBBx5wtt12221avny5ZsyYoTvvvFN5eXl65plntHPnTn366acuMe/Zs0cTJ07UbbfdplmzZqlXr17kS4AZDADNzrJlywxJZ3xceOGFLq/p2rWrMW3aNOfzefPmGZKMzz77zNlWWFhoWK1WQ5KRl5fnbGvVqpVxzTXXGBUVFc6+v/vd7wxJLvt85JFHjLZt2xr//ve/XY59//33G4GBgcaBAwcMwzCMvLw8Q5IRHh5uFBUVOfutWbPGkGSsXbv2jO9/48aNhiRj9erVxpEjR4wOHToY1113nXP7tGnTjLZt2zqfl5SUGO3btzdmzZrlsp+CggLDarW6tA8ZMsQYMmRItWNOmzbN6Nq1q/N51XsIDQ01CgsLXfr269fPiIiIMA4fPuxs++KLL4yAgABj6tSpzrbk5GRDknHzzTe7vP5Xv/qVER4e7nz+1FNPGZKMH3/88YzjAgAAGo4863951pkcP37c5fmJEyeMmJgYY9iwYc62I0eOGMHBwcZ9993n0vfOO+802rZtaxw9etQwDMP417/+ZUgyXnvtNZd+6enp1dq7du1qSDLS09Nd+pIvAd7HJXtAM/bss89q/fr11R59+/Y962vff/99XX755YqLi3O2nXvuuZo8ebJLv48++kgnTpzQHXfcIYvF4myfN29etX2uXr1aV155pTp06KD//ve/zseIESNUXl6uzZs3u/S/4YYb1KFDB+fzK6+8UpL0n//8p07vX5KsVqvmzZunf/7zn9q5c2eNfdavX68jR45o4sSJLnEFBgZqwIAB2rhxY52Pd7oJEybo3HPPdT7Pz89Xdna2pk+frrCwMGd73759NXLkSL3//vvV9nH77be7PL/yyit1+PBhFRcXS5Lat28vSVqzZk2NU/0BAIDnkWed2amzvX766Sc5HA5deeWVLmt7Wq1WjR07Vq+//roMw5AklZeX64033tC4cePUtm1b53uzWq0aOXKky3vr37+/2rVrVy1Xi4qKcrncTyJfAszAJXtAMxYXF6dLL720WntVonIm3377rQYMGFCtvVevXtX6SVLPnj1d2s8991yXJEeSvvnmG+3atculQHOqwsJCl+ddunSpFrekamsFnM1dd92lp556SgsXLtSaNWuqbf/mm28kScOGDavx9aGhoW4d71RRUVEuz6vG6/RxlKQ+ffpo3bp1OnbsmDMBk848DqGhobrhhhv08ssv65ZbbtH999+v4cOHa/z48fr1r3/NHX0AAGgk5Fln9u677+rRRx9Vdna2ysrKnO2nFtYkaerUqXrjjTf0r3/9S4MHD9ZHH32kQ4cO6aabbnL2+eabb+RwOBQREVHjsU5/b6fnX5LIlwATUJAC4DMqKio0cuRI3XvvvTVuv+CCC1yeBwYG1tiv6he0uqqaJbVw4cIaZ0lV/Ur26quvymazVdveosX//lNqsVhqPH55eXmNx65pLQh3nW0cWrdurc2bN2vjxo167733lJ6erjfeeEPDhg3Thx9+WOvrAQCA/zArz6rJv/71L1133XUaPHiwnnvuOdntdrVs2VLLli3TypUrXfqOGjVKHTt21IoVKzR48GCtWLFCNptNI0aMcPapqKhQRESEXnvttRqPd3oRrqb8i3wJ8D4KUgDqpWvXrs6ZQ6fas2dPtX5S5S9X559/vrP9xx9/rPYLW/fu3XX06FGXBMNb5s2bpyVLliglJcU5ZfvUuCQpIiLirLF16NChxqnsVb9gnk3VeJ0+jpL09ddf65xzznGZHVVXAQEBGj58uIYPH64//elPevzxx/X73/9eGzduNGW8AQBA7fwtzzrdP/7xDwUHB2vdunUKCgpyti9btqxa38DAQE2aNEnLly/X4sWL9c4772jWrFkuBaLu3bvro48+0sCBAxv0Yx/5EuBdzD0EUC9jxozRtm3blJmZ6Wz78ccfq/0yNWLECLVs2VJ//vOfXX5RO/1OLlLl3eK2bt2qdevWVdt25MgR/fLLL557A6epmiW1Zs0aZWdnu2wbNWqUQkND9fjjj9d4Z5wff/zR+e/u3bvr66+/dmn74osv6nyHFrvdrn79+umvf/2rjhw54mzPycnRhx9+qDFjxrj3xiQVFRVVa+vXr58kuUyRBwAAvsHf8qzTBQYGymKxuMwg379/v955550a+99000366aefdNttt+no0aOaMmWKy/brr79e5eXleuSRR6q99pdffnHJqWpDvgR4HzOkANTLvffeq1dffVUJCQm66667nLcj7tq1q3bt2uXsd+655+ruu+9Wamqqrr32Wo0ZM0Y7d+7UBx98oHPOOcdln/fcc4/++c9/6tprr9X06dPVv39/HTt2TLt379abb76p/fv3V3uNJ1WtJfXFF1+4zEIKDQ1VWlqabrrpJsXGxurGG2/UueeeqwMHDui9997TwIED9cwzz0iSbr75Zv3pT3/SqFGjNHPmTBUWFur555/XhRde6Fxk/GyefPJJjR49WvHx8Zo5c6Z+/vln/fnPf5bVatXChQvdfl8PP/ywNm/erGuuuUZdu3ZVYWGhnnvuOXXu3FmDBg1ye38AAKBx+UOe9Y9//ENff/11tfZp06bpmmuu0Z/+9CclJCRo0qRJKiws1LPPPqsePXq4vL8ql1xyiWJiYrR69Wr16dNHsbGxLtuHDBmi2267TampqcrOztbVV1+tli1b6ptvvtHq1au1dOlS/frXvz5jvORLgPdRkAJQL3a7XRs3btQdd9yhRYsWKTw8XLfffrs6deqkmTNnuvR99NFHFRwcrOeff14bN27UgAED9OGHH+qaa65x6demTRtlZGTo8ccf1+rVq/W3v/1NoaGhuuCCC5SSkiKr1dqo76l9+/aaN2+eUlJSqm2bNGmSOnXqpEWLFunJJ59UWVmZzjvvPF155ZWaMWOGs1+fPn30t7/9TQ899JAWLFig6Ohovfrqq1q5cqU2bdpUpzhGjBih9PR0JScn66GHHlLLli01ZMgQLV68uMZFOM/muuuu0/79+/XKK6/ov//9r8455xwNGTLEK2MKAADc5w951qpVq2psHzp0qIYNG6a//OUvWrRokebNm6eoqCgtXrxY+/fvr7EgJVUubn7vvfe6LGZ+queff179+/fXCy+8oN/97ndq0aKFunXrpilTpmjgwIFnjZd8CfA+i+GJVekAAAAAAGgkS5cu1fz587V///5qdwAE0DRRkAIAAAAA+CzDMHTxxRcrPDxcGzduNDscAB7CJXsAAAAAAJ9z7Ngx/fOf/9TGjRu1e/durVmzxuyQAHgQM6QAAAAAAD5n//79ioqKUvv27fWb3/xGjz32mNkhAfAgClIAAAAAAADwqgCzAwAAAAAAAEDzQkEKAAAAAAAAXuX3i5pXVFTohx9+UEhIiCwWi9nhAAAAH2YYhkpKStSpUycFBPC7nUQuBQAA6sbdPMrvC1I//PCDIiMjzQ4DAAA0IQcPHlTnzp3NDsMnkEsBAAB31DWP8vuCVEhIiKTKAQkNDTU5GgAA4MuKi4sVGRnpzB9ALgUAAOrG3TzK7wtSVVPLQ0NDSaIAAECdcGna/5BLAQAAd9Q1j2JxBAAAAAAAAHgVBSkAAAAAAAB4FQUpAAAAAAAAeBUFKQAAAAAAAHgVBSkAAAAAAAB4FQUpAAAAAAAAeBUFKQAAAAAAAHgVBSkAAAAAAAB4VQuzAwAAAP6vvMJQZl6RCktKFRESrLioMAUGWMwOCybinPAcxrJSQ8fBE+PoCzE0FOPgOzH4Al8YB3+Ige9VzUwtSC1cuFApKSkubb169dLXX38tSSotLdVvf/tbrVq1SmVlZRo1apSee+45dezY0YxwAQBAPaTn5Ctlba7yHaXONrs1WMmJ0UqIsZsYWdNytrypJqtXr9aDDz6o/fv3q2fPnlq8eLHGjBnT2KGeFeeE5zCWlRo6Dp4YR1+IoaEYB9+JwRf4wjj4Qwx8r2pn+iV7F154ofLz852PTz75xLlt/vz5Wrt2rVavXq2MjAz98MMPGj9+vInRAgAAd6Tn5Gv2iiyXBEiSChylmr0iS+k5+SZF1jSdKW863ZYtWzRx4kTNnDlTO3fu1Lhx4zRu3Djl5OR4MeLqOCc8h7Gs1NBx8MQ4+kIMDcU4+E4MvsAXxsEfYuB7dWamF6RatGghm83mfJxzzjmSJIfDob/85S/605/+pGHDhql///5atmyZtmzZom3btpkcNQAAOJvyCkMpa3Nl1LCtqi1lba7KK2rqgZrUljfVZOnSpUpISNA999yjPn366JFHHlFsbKyeeeYZL0bsinPCcxjLSg0dB0+Moy/E0FCMg+/E4At8YRz8IQa+V2dnekHqm2++UadOnXT++edr8uTJOnDggCRpx44dOnnypEaMGOHs27t3b3Xp0kVbt26tdX9lZWUqLi52eQAAAO/LzCuq9mvcqQxJ+Y5SZeYVeS+oJq62vKkmW7dudcmjJGnUqFFnzKOkxs2lOCc8h7Gs1NBx8MQ4+kIMDcU4+E4MvsAXxsEfYuB7dXamFqQGDBig5cuXKz09XWlpacrLy9OVV16pkpISFRQUqFWrVmrfvr3Lazp27KiCgoJa95mamiqr1ep8REZGNvK7AAAANSksqT0Bqk+/5u5MeVNNCgoKqq27ebY8SmrcXIpzwnMYy0oNHQdPjKMvxNBQjIPvxOALfGEc/CEGvldnZ+qi5qNHj3b+u2/fvhowYIC6du2qv//972rdunW99pmUlKQFCxY4nxcXF1OUAgDABBEhwR7t19ydKW+aOXOmx47TmLkU54TnMJaVGjoOnhhHX4ihoRgH34nBF/jCOPhDDHyvzs70S/ZO1b59e11wwQXau3evbDabTpw4oSNHjrj0OXTokGw2W637CAoKUmhoqMsDAAB4X1xUmOzWYNV2Q2GLKu/wEhcV5s2w/MapeVNNbDabDh065NJ2tjxKatxcinPCcxjLSg0dB0+Moy/E0FCMg+/E4At8YRz8IQa+V2fnUwWpo0ePat++fbLb7erfv79atmypjz/+2Ll9z549OnDggOLj402M8n/KKwxt3XdYa7K/19Z9h/1+cTsAANwRGGBRcmK0JFVLhKqeJydGKzCgtjQJZ3Jq3lST+Ph4lzxKktavX29qHsU54TmMZaWGjoMnxtEXYmgoxsF3YvAFvjAO/hAD36uzM7UgdffddysjI0P79+/Xli1b9Ktf/UqBgYGaOHGirFarZs6cqQULFmjjxo3asWOHZsyYofj4eF1++eVmhi2p8taJgxZv0MSXtumuVdma+NI2DVq8odncBhQAgLpIiLErbUqsbFbXqeA2a7DSpsQqIabmYgqqO1PeJElTp05VUlKSs/9dd92l9PR0/fGPf9TXX3+thQsX6vPPP9fcuXPNeguSOCc8ibGs1NBx8MQ4+kIMDcU4+E4MvsAXxsEfYuB7dWYWwzBMm9Zz4403avPmzTp8+LDOPfdcDRo0SI899pi6d+8uSSotLdVvf/tbvf766yorK9OoUaP03HPPnXWq+amKi4tltVrlcDg8NuU8PSdfs1dkVbt1YlVN0ewPFQAAX1NeYSgzr0iFJaWKCKmcGu6LvzA3Rt7gKWfLm4YOHapu3bpp+fLlztesXr1aDzzwgPbv36+ePXvqiSee0JgxY9w6bmONSVM5J5oCxrJSQ8fBE+PoCzE0FOPgOzH4Al8YB3+Iobl8r9zNGUwtSHmDp5Oo8gpDgxZvqPXWiRZVVho/uW9Ys/wPFgAATZkvF6TMwpgAAIC6cDdn8Kk1pJqCzLyiWotRkmRIyneUKjOvyHtBAQAAAAAANCEUpNxUWFJ7Mao+/QAAAAAAAJobClJuiggJPnsnN/oBAAAAAAA0NxSk3BQXFSa7NbjaLROrWCTZrZULhAEAAAAAAKA6ClJuCgywKDkxWpKqFaWqnicnRrOgOQAAAAAAQC0oSNVDQoxdaVNiZbO6XpZnswYrbUqsEmLsJkUGAAAAAADg+1qYHUBTlRBj18homzLzilRYUqqIkMrL9JgZBQAAAAAAcGYUpBogMMCi+O7hZocBAAAAAADQpHDJHgAAAAAAALyKghQAAAAAAAC8ioIUAAAAAAAAvIqCFAAAAAAAALyKghQAAAAAAAC8ioIUAAAAAAAAvIqCFAAAAAAAALyKghQAAAAAAAC8ioIUAAAAAAAAvIqCFAAAAAAAALyKghQAAAAAAAC8qoXZAQC+oLzCUGZekQpLShUREqy4qDAFBliaXQwAAAAAAHgDBSk0e+k5+UpZm6t8R6mzzW4NVnJitBJi7M0mBgAAAAAAvIVL9tCspefka/aKLJdCkCQVOEo1e0WW0nPym0UMAAAAAAB4EwUpNFvlFYZS1ubKqGFbVVvK2lyVV9TUw39iAAAAAADA2yhIodnKzCuqNivpVIakfEepMvOK/DoGAAAAAAC8jYIUmq3CktoLQfXp11RjAAAAAADA2yhIodmKCAn2aL+mGgMAAAAAAN5GQQrNVlxUmOzWYFlq2W5R5Z3u4qLC/DoGAAAAAAC8jYIUGqy8wtDWfYe1Jvt7bd13uMkswB0YYFFyYrQkVSsIVT1PToxWYEBt5SL/iAEAAAAAAG/zmYLUokWLZLFYNG/ePGfb0KFDZbFYXB633367eUGimvScfA1avEETX9qmu1Zla+JL2zRo8Qal5+SbHVqdJMTYlTYlVjar6yVxNmuw0qbEKiHG3ixiAAAAAADAm1qYHYAkbd++XS+88IL69u1bbdusWbP08MMPO5+3adPGm6HhDNJz8jV7RZZOnw9V4CjV7BVZTaaYkhBj18homzLzilRYUqqIkMpL5Lw5K8kXYgAAAAAAwFtML0gdPXpUkydP1ksvvaRHH3202vY2bdrIZrOZEBnOpLzCUMra3GrFKEkyVHm5WcraXI2MtjWJokpggEXx3cObfQwAAAAAAHiD6ZfszZkzR9dcc41GjBhR4/bXXntN55xzjmJiYpSUlKTjx4+fcX9lZWUqLi52ecDzMvOKlO8orXW7ISnfUarMvCLvBQUAAAAAAJoEU2dIrVq1SllZWdq+fXuN2ydNmqSuXbuqU6dO2rVrl+677z7t2bNHb731Vq37TE1NVUpKSmOFjP+vsKT2YlR9+gEAAAAAgObDtILUwYMHddddd2n9+vUKDg6usc+tt97q/PdFF10ku92u4cOHa9++ferevXuNr0lKStKCBQucz4uLixUZGenZ4KGIkJo/s/r2AwAAAAAAzYdpBakdO3aosLBQsbGxzrby8nJt3rxZzzzzjMrKyhQYGOjymgEDBkiS9u7dW2tBKigoSEFBQY0XOCRJcVFhsluDVeAorXEdKYsq7xIXFxXm7dAAAAAAAICPM60gNXz4cO3evdulbcaMGerdu7fuu+++asUoScrOzpYk2e2+f+c2fxcYYFFyYrRmr8iSRXIpSlUtYZ6cGN0kFjT3J+UVBnfqAwAAAAD4PNMKUiEhIYqJiXFpa9u2rcLDwxUTE6N9+/Zp5cqVGjNmjMLDw7Vr1y7Nnz9fgwcPVt++fU2KGqdKiLErbUqsUtbmuixwbrMGKzkxWgkxFA69KT0nv9pnYeezAAAAAAD4IFMXNT+TVq1a6aOPPtKSJUt07NgxRUZGasKECXrggQfMDg2nSIixa2S0jVk5JkvPydfsFVnVLp8scJRq9oospU2JpSgFAAAAAPAZPlWQ2rRpk/PfkZGRysjIMC8Y1FlggEXx3cPNDqPZKq8wlLI2t8a1vAxVXkKZsjZXI6NtFAoBAAAAAD4hwOwAADRMZl6Ry2V6pzMk5TtKlZlX5L2gAAAAAAA4AwpSQBNXWFJ7Mao+/QAAAAAAaGwUpIAmLiIk2KP9AABNw6JFi2SxWDRv3rxa+yxfvlwWi8XlERzM/w8AAID5fGoNKQDui4sKk90arAJHaY3rSFlUeefDuKgwb4cGAGgk27dv1wsvvFCnOw+HhoZqz549zucWC+sJAgAA8zFDCmjiAgMsSk6MllRZfDpV1fPkxGgWNAcAP3H06FFNnjxZL730kjp06HDW/haLRTabzfno2LGjF6IEAAA4MwpSgB9IiLErbUqsbFbXyzBs1mClTYlVQozdpMgAAJ42Z84cXXPNNRoxYkSd+h89elRdu3ZVZGSkxo4dqy+//PKM/cvKylRcXOzyAAAA8DQu2QP8REKMXSOjbcrMK1JhSakiQiov02NmFAD4j1WrVikrK0vbt2+vU/9evXrplVdeUd++feVwOPSHP/xBV1xxhb788kt17ty5xtekpqYqJSXFk2EDAABUYzEMo6ZlZ/xGcXGxrFarHA6HQkNDzQ4HAAD4MF/OGw4ePKhLL71U69evd64dNXToUPXr109Lliyp0z5OnjypPn36aOLEiXrkkUdq7FNWVqaysjLn8+LiYkVGRvrkmAAAAN/hbh7FDCkAAIAmYMeOHSosLFRsbKyzrby8XJs3b9YzzzyjsrIyBQYGnnEfLVu21CWXXKK9e/fW2icoKEhBQUEeixsAAKAmFKQAAACagOHDh2v37t0ubTNmzFDv3r113333nbUYJVUWsHbv3q0xY8Y0VpgAAAB1QkEKAACgCQgJCVFMTIxLW9u2bRUeHu5snzp1qs477zylpqZKkh5++GFdfvnl6tGjh44cOaInn3xS3377rW655Ravxw8AAHAqClIAAAB+4sCBAwoI+N9NlH/66SfNmjVLBQUF6tChg/r3768tW7YoOjraxCgBAABY1BwAAMCJvKE6xgQAANSFuzlDwFl7AAAAAAAAAB5EQQoAAAAAAABeRUEKAAAAAAAAXkVBCgAAAAAAAF5FQQoAAAAAAABeRUEKAAAAAAAAXtXC7AAA+I7yCkOZeUUqLClVREiw4qLCFBhgMTssAAAAAICfoSAFQJKUnpOvlLW5yneUOtvs1mAlJ0YrIcZuYmQAAAAAAH/DJXsAlJ6Tr9krslyKUZJU4CjV7BVZSs/JNykyAAAAAIA/oiAFNHPlFYZS1ubKqGFbVVvK2lyVV9TUAwAAAAAA91GQApq5zLyiajOjTmVIyneUKjOvyHtBAQAAAAD8GgUpoJkrLKm9GFWffgAAAAAAnA0FKaCZiwgJ9mg/AAAAAADOhoIU0MzFRYXJbg2WpZbtFlXebS8uKsybYQEAAAAA/BgFKaCZCwywKDkxWpKqFaWqnicnRiswoLaSFQAAAAAA7vGZgtSiRYtksVg0b948Z1tpaanmzJmj8PBwtWvXThMmTNChQ4fMCxLwUwkxdqVNiZXN6npZns0arLQpsUqIsZsUGQAAAADAH7UwOwBJ2r59u1544QX17dvXpX3+/Pl67733tHr1almtVs2dO1fjx4/Xp59+alKkgP9KiLFrZLRNmXlFKiwpVURI5WV6zIwCAAAAAHia6QWpo0ePavLkyXrppZf06KOPOtsdDof+8pe/aOXKlRo2bJgkadmyZerTp4+2bdumyy+/3KyQAb8VGGBRfPdws8MAAAAAAPg50y/ZmzNnjq655hqNGDHCpX3Hjh06efKkS3vv3r3VpUsXbd26tdb9lZWVqbi42OUBAAAAAAAA32HqDKlVq1YpKytL27dvr7atoKBArVq1Uvv27V3aO3bsqIKCglr3mZqaqpSUFE+HCgAAAAAAAA8xbYbUwYMHddddd+m1115TcHDw2V9QR0lJSXI4HM7HwYMHPbZvAAAAAAAANJxpBakdO3aosLBQsbGxatGihVq0aKGMjAw9/fTTatGihTp27KgTJ07oyJEjLq87dOiQbDZbrfsNCgpSaGioywMAAAAAAAC+w7RL9oYPH67du3e7tM2YMUO9e/fWfffdp8jISLVs2VIff/yxJkyYIEnas2ePDhw4oPj4eDNCBgAAAAAAgAeYVpAKCQlRTEyMS1vbtm0VHh7ubJ85c6YWLFigsLAwhYaG6o477lB8fDx32AMAAAAAAGjCTF3U/GyeeuopBQQEaMKECSorK9OoUaP03HPPmR0WAAAAAAAAGsBiGIZhdhCNqbi4WFarVQ6Hg/WkAADAGZE3VMeYAACAunA3ZzBtUXMAAAAAAAA0TxSkAAAAAAAA4FUUpAAAAAAAAOBVFKQAAAAAAADgVT59lz0AMEN5haHMvCIVlpQqIiRYcVFhCgywmB0W6sETnyXnAwAAAOB5FKQA4BTpOflKWZurfEeps81uDVZyYrQSYuwmRgZ3eeKz5HwAAAAAGgeX7AHA/5eek6/ZK7Jcig+SVOAo1ewVWUrPyTcpMrjLE58l5wMAAADQeChIAYAqL8tKWZsro4ZtVW0pa3NVXlFTD/gST3yWnA8AAABA46IgBQCSMvOKqs2EOZUhKd9Rqsy8Iu8FhXrxxGfJ+QAAAAA0LrcLUgcPHtR3333nfJ6Zmal58+bpxRdf9GhgAOBNhSW1Fx/q0w/m8cRnyfkAAAAANC63C1KTJk3Sxo0bJUkFBQUaOXKkMjMz9fvf/14PP/ywxwMEAG+ICAn2aD+YxxOfJecDAAAA0LjcLkjl5OQoLi5OkvT3v/9dMTEx2rJli1577TUtX77c0/EBgFfERYXJbg2WpZbtFlXeXS0uKsybYaEePPFZcj4AAAAAjcvtgtTJkycVFBQkSfroo4903XXXSZJ69+6t/HzuOASgaQoMsCg5MVqSqhUhqp4nJ0YrMKC2EgV8hSc+S84HAAAAoHG5XZC68MIL9fzzz+tf//qX1q9fr4SEBEnSDz/8oPDwcI8HCADekhBjV9qUWNmsrpdh2azBSpsSq4QYu0mRwV2e+Cw5HwAAAIDGYzEMw617Vm/atEm/+tWvVFxcrGnTpumVV16RJP3ud7/T119/rbfeeqtRAq2v4uJiWa1WORwOhYaGmh0OgCagvMJQZl6RCktKFRFSeVkWM2GaJk98lpwPzQt5Q3WMyZk19L8RvvDfGGKAP/KFc8rsGMw+vq9gHLzH3ZzB7YKUJJWXl6u4uFgdOnRwtu3fv19t2rRRRESEu7trVCRRAACgrryRNxw5ckTt27dv8H4WLVqkpKQk3XXXXVqyZEmt/VavXq0HH3xQ+/fvV8+ePbV48WKNGTOmzschl6pdek6+UtbmKt/xvztu2q3BSk6MrtMsyoa+3hOIAf7IF84ps2Mw+/i+gnHwLndzBrcv2ZOkwMBAl2KUJHXr1s3nilEAAABmWrx4sd544w3n8+uvv17h4eE677zz9MUXX9R7v9u3b9cLL7ygvn37nrHfli1bNHHiRM2cOVM7d+7UuHHjNG7cOOXk5NT72KiUnpOv2SuyXP7IkaQCR6lmr8hSes6Z11Zt6Os9gRjgj3zhnDI7BrOP7ysYB9/ndkHq0KFDuummm9SpUye1aNFCgYGBLg8AAABUev755xUZGSlJWr9+vdavX68PPvhAo0eP1j333FOvfR49elSTJ0/WSy+9VO0HwtMtXbpUCQkJuueee9SnTx898sgjio2N1TPPPFOvY6NSeYWhlLW5qukyg6q2lLW5Kq+o+UKEhr7eE4gB/sgXzimzYzD7+L6CcWgaWrj7gunTp+vAgQN68MEHZbfbZbFw7SUAAEBNCgoKnAWpd999V9dff72uvvpqdevWTQMGDKjXPufMmaNrrrlGI0aM0KOPPnrGvlu3btWCBQtc2kaNGqV33nmn1teUlZWprKzM+by4uLhecfqzzLyiar+4n8qQlO8oVWZekeK7V7/pT0Nf7wnEAH/kC+eU2TGYfXxfwTg0DW4XpD755BP961//Ur9+/RohHAAAAP/RoUMHHTx4UJGRkUpPT3cWkAzDUHl5udv7W7VqlbKysrR9+/Y69S8oKFDHjh1d2jp27KiCgoJaX5OamqqUlBS3Y2tOCktq/yOnLv0a+npPIAb4I184p8yOwezj+wrGoWlw+5K9yMhI1WMddAAAgGZn/PjxmjRpkkaOHKnDhw9r9OjRkqSdO3eqR48ebu3r4MGDuuuuu/Taa68pODi4McKVJCUlJcnhcDgfBw8ebLRjNVURIXUb/9r6NfT1nkAM8Ee+cE6ZHYPZx/cVjEPT4HZBasmSJbr//vu1f//+RggHAADAfzz11FO64447FB0drfXr16tdu3aSpPz8fP3mN79xa187duxQYWGhYmNj1aJFC7Vo0UIZGRl6+umn1aJFixpnXNlsNh06dMil7dChQ7LZbLUeJygoSKGhoS4PuIqLCpPdGqzaFq6wqPIuTnFRYY3yek8gBvgjXzinzI7B7OP7CsahaXC7IHXDDTdo06ZN6t69u0JCQhQWFubyAAAAgHTy5EnddtttGj9+vJYuXapLLrnEuW3+/Pm65ZZb3Nrf8OHDtXv3bmVnZzsfl156qSZPnqzs7Owaby4THx+vjz/+2KVt/fr1io+Pr9+bgiQpMMCi5MRoSar2x07V8+TEaAUG1PynUENf7wnEAH/kC+eU2TGYfXxfwTg0DRbDzevv/vrXv55x+7Rp0xoUkKcVFxfLarXK4XDwCx8AADgjT+cNVqtV2dnZioqK8kB01Q0dOlT9+vXTkiVLJElTp07Veeedp9TUVEnSli1bNGTIEC1atEjXXHONVq1apccff1xZWVmKiYmp0zHIpWqXnpOvlLW5Lgvn2q3BSk6MVkKMvdFf7wnEAH/kC+eU2TGYfXxfwTh4l7s5g9sFqaaGJAoAANSVp/OGadOmqV+/fpo/f74Hoqvu9ILU0KFD1a1bNy1fvtzZZ/Xq1XrggQe0f/9+9ezZU0888YTGjBlT52OQS51ZeYWhzLwiFZaUKiKk8vIPd35xb+jrPYEY4I984ZwyOwazj+8rGAfv8UpBqry8XO+8846++uorSdKFF16o6667rsap4mYjiQIAAHXl6bzh0Ucf1R//+EcNHz5c/fv3V9u2bV2233nnnQ0+RmMjlwIAAHXR6AWpvXv3asyYMfr+++/Vq1cvSdKePXsUGRmp9957T927d69f5I2EJAoAANSVp/OGM12qZ7FY9J///KfBx2hs5FIAAKAu3M0ZWrh7gDvvvFPdu3fXtm3bnIuYHz58WFOmTNGdd96p9957z/2oUW9MPwQAwHfl5eWZHQIAAIBPcvsuexkZGXriiSdc7qgXHh6uRYsWKSMjw619paWlqW/fvs5bCsfHx+uDDz5wbh86dKgsFovL4/bbb3c3ZL+VnpOvQYs3aOJL23TXqmxNfGmbBi3eoPScfLNDAwAApzhx4oT27NmjX375xexQAAAAfILbBamgoCCVlJRUaz969KhatWrl1r46d+6sRYsWaceOHfr88881bNgwjR07Vl9++aWzz6xZs5Sfn+98PPHEE+6G7JfSc/I1e0WWy90CJKnAUarZK7IoSgEA4AOOHz+umTNnqk2bNrrwwgt14MABSdIdd9yhRYsWmRwdAACAedwuSF177bW69dZb9dlnn8kwDBmGoW3btun222/Xdddd59a+EhMTNWbMGPXs2VMXXHCBHnvsMbVr107btm1z9mnTpo1sNpvzwdoFlZfppazNVU2Lf1W1pazNVXmFX99AEQAAn5eUlKQvvvhCmzZtUnBwsLN9xIgReuONN0yMDAAAwFxuF6Sefvppde/eXfHx8QoODlZwcLAGDhyoHj16aOnSpfUOpLy8XKtWrdKxY8cUHx/vbH/ttdd0zjnnKCYmRklJSTp+/PgZ91NWVqbi4mKXh7/JzCuqNjPqVIakfEepMvOKvBcUAACo5p133tEzzzyjQYMGyWL53xqPF154ofbt22diZAAAAOZya1FzwzBUXFysVatW6fvvv9dXX30lSerTp4969OhRrwB2796t+Ph4lZaWql27dnr77bcVHR0tSZo0aZK6du2qTp06adeuXbrvvvu0Z88evfXWW7XuLzU1VSkpKfWKpakoLKm9GFWffgAAoHH8+OOPioiIqNZ+7NgxlwIVAABAc+N2QapHjx768ssv1bNnz3oXoU7Vq1cvZWdny+Fw6M0339S0adOUkZGh6Oho3Xrrrc5+F110kex2u4YPH659+/ape/fuNe4vKSlJCxYscD4vLi5WZGRkg+P0JREhwWfv5EY/AADQOC699FK99957uuOOOyTJWYR6+eWXXWaEAwAANDduFaQCAgLUs2dPHT58WD179vRIAK1atXIWtvr376/t27dr6dKleuGFF6r1HTBggCRp7969tRakgoKCFBQU5JHYfFVcVJjs1mAVOEprXEfKIslmDVZcVFgNWwEAgLc8/vjjGj16tHJzc/XLL79o6dKlys3N1ZYtW9y+OzEAAIA/cXsNqUWLFumee+5RTk5OY8SjiooKlZWV1bgtOztbkmS32xvl2E1FYIBFyYmVlzWePtm/6nlyYrQCA7gUAAAAMw0aNEjZ2dn65ZdfdNFFF+nDDz9URESEtm7dqv79+5sdHgAAgGkshmG4dSu2Dh066Pjx4/rll1/UqlUrtW7d2mV7UVHdF9JOSkrS6NGj1aVLF5WUlGjlypVavHix1q1bp/PPP18rV67UmDFjFB4erl27dmn+/Pnq3LmzW78oFhcXy2q1yuFw+N0d+tJz8pWyNtdlgXO7NVjJidFKiGneRTsAAOrDn/OG+mJMAABAXbibM7h1yZ4kLVmypD5x1aiwsFBTp05Vfn6+rFar+vbtq3Xr1mnkyJE6ePCgPvroIy1ZskTHjh1TZGSkJkyYoAceeMBjx2/qEmLsGhltU2ZekQpLShURUnmZHjOjAADwDUOGDNHMmTP1f//3f9V+xAMAAGjO3J4h1dTwqx4AAKgrT+cN8+bN08qVK1VWVqbrr79eM2fO1OWXX+6BSL2HXAoAANSFuzmD22tIHThw4IwPAAAAVFqyZIl++OEHLVu2TIWFhRo8eLCio6P1hz/8QYcOHTI7PAAAANO4PUMqICDAecvimpSXlzc4KE/iVz0AAFBXjZ03FBYW6sUXX9Rjjz2m8vJyjRkzRnfeeaeGDRvm8WN5CrkUAACoi0ZfQ2rnzp0uz0+ePKmdO3fqT3/6kx577DF3dwcAANAsZGZmatmyZVq1apUiIiI0ffp0ff/997r22mv1m9/8Rn/4wx/MDhEAAMBr3C5IXXzxxdXaLr30UnXq1ElPPvmkxo8f75HAAAAAmrrCwkK9+uqrWrZsmb755hslJibq9ddf16hRo5wzzqdPn66EhAQKUgAAoFlxuyBVm169emn79u2e2l2zUF5hcIc8oBHw3WIMAF/RuXNnde/eXTfffLOmT5+uc889t1qfvn376rLLLjMhOgAAAPO4XZAqLi52eW4YhvLz87Vw4UL17NnTY4H5u/ScfKWszVW+o9TZZrcGKzkxWgkxdhMjA5o2vluMAeBLPv74Y1155ZVn7BMaGqqNGzd6KSIAAADf4JFFzQ3DUGRkpFatWqX4+HiPBthQvrgQZ3pOvmavyNLpA181qmlTYvmjEagHvluMAdBQvpg3mI0xAQAAddHoi5qf/gteQECAzj33XPXo0UMtWnjsCkC/VV5hKGVtbrU/FiXJUOUfjSlrczUy2sblNYAb+G4xBoCvevPNN/X3v/9dBw4c0IkTJ1y2ZWVlmRQVAACAuQLcfcGQIUNcHldeeaV69+5NMaqOMvOKXC6jOZ0hKd9Rqsy8Iu8FBfgBvluMAeCLnn76ac2YMUMdO3bUzp07FRcXp/DwcP3nP//R6NGjzQ4PAADANG4XpCTp1Vdf1cCBA9WpUyd9++23kqSnnnpKa9as8Whw/qiwpPY/FuvTD0AlvluMAeCLnnvuOb344ov685//rFatWunee+/V+vXrdeedd8rhcJgdHgAAgGncLkilpaVpwYIFGjNmjI4cOaLy8nJJUocOHbRkyRJPx+d3IkKCPdoPQCW+W4wB4IsOHDigK664QpLUunVrlZSUSJJuuukmvf7662aGBgAAYCq3C1J//vOf9dJLL+n3v/+9AgMDne2XXnqpdu/e7dHg/FFcVJjs1mDVtnqLRZV3w4qLCvNmWECTx3eLMQB8kc1mU1FR5WWyXbp00bZt2yRJeXl5cvO+MgAAAH7F7YJUXl6eLrnkkmrtQUFBOnbsmEeC8meBARYlJ0ZLUrU/GqueJydGs+Aw4Ca+W4wB/F95haGt+w5rTfb32rrvsMorfL+gM2zYMP3zn/+UJM2YMUPz58/XyJEjdcMNN+hXv/qVydEBAACYx+2VyKOiopSdna2uXbu6tKenp6tPnz4eC8yfJcTYlTYlVilrc10WILZZg5WcGM0t2YF64rvFGMB/pefkVzuv7U3gvH7xxRdVUVEhSZozZ47Cw8O1ZcsWXXfddbr99ttNjg4AAMA8FsPN+eIvv/yyFi5cqD/+8Y+aOXOmXn75Ze3bt0+pqal6+eWXdeONNzZWrPVSXFwsq9Uqh8Oh0NBQs8NxUV5hKDOvSIUlpYoIqbyMhpkLQMPx3WIM4F/Sc/I1e0WWTk9Yqs7otCmxHitKeStvOHLkiN5//31NmjSp0Y7hKb6cSwEAAN/hbs7gdkFKkl577TUtXLhQ+/btkyR16tRJKSkpmjlzpvsRNzKSKAAAmq7yCkODFm9wmRl1KosqZwB+ct8wjxRdvZU3fPHFF4qNjXXeHMaXkUsBAIC6cDdncPuSPUmaPHmyJk+erOPHj+vo0aOKiIioz24AAADOKDOvqNZilCQZkvIdpcrMK1J893DvBQYAAIAGqVdBqkqbNm3Upk0bT8UCAADgorCk9mJUffoBAADAN9S5IHXVVVfJYjnzVHiLxaKPP/64wUEBAABIUkRIsEf7AQAAwDfUuSDVr1+/WreVlJRo5cqVKisr80RMAAAAkqS4qDDZrcEqcJRWW9Rc+t8aUnFRYd4O7YyefvrpM27//vvvvRQJAACAb6pzQeqpp56q1vbLL7/o2Wef1WOPPabzzjtPjzzyiEeDAwAAzVtggEXJidGavSJLFsmlKFU1bzs5Mdrn7iJZU950ui5dunghEgAAAN9U7zWkXnvtNT300EP6+eeftXDhQt16661q0aJBS1IBAABUkxBjV9qUWKWszXVZ4NxmDVZyYrQSYuwmRlezvLw8s0MAAADwaW5XkNLT03X//fcrLy9Pd999txYsWKC2bds2RmwAAACSKotSI6NtyswrUmFJqSJCKi/T87WZUQAAAKibOhekMjMzdd9992nbtm26/fbb9dFHH+mcc85pzNgAAACcAgMsiu8ebnYYAAAA8IA6F6Quv/xytW7dWrfffruioqK0cuXKGvvdeeedHgsOAAAAAAAA/qfOBakuXbrIYrHonXfeqbWPxWKhIAUAAAAAAIAzqnNBav/+/Y0YBgAAAAAAAJoLbosHAGgU5RUGC1ADkioqKrR3714VFhaqoqLCZdvgwYNNigoAAMBcAWYePC0tTX379lVoaKhCQ0MVHx+vDz74wLm9tLRUc+bMUXh4uNq1a6cJEybo0KFDJkYMAKiL9Jx8DVq8QRNf2qa7VmVr4kvbNGjxBqXn5JsdGuBV27ZtU48ePdSnTx8NHjxYQ4cOdT6uuuoqt/Z1trzpdMuXL5fFYnF5BAcHN/QtAQAAeISpBanOnTtr0aJF2rFjhz7//HMNGzZMY8eO1ZdffilJmj9/vtauXavVq1crIyNDP/zwg8aPH29myACAs0jPydfsFVnKd5S6tBc4SjV7RRZFKTQrt99+uy699FLl5OSoqKhIP/30k/NRVFTk1r7OljfVJDQ0VPn5+c7Ht99+29C3BAAA4BEWwzAMs4M4VVhYmJ588kn9+te/1rnnnquVK1fq17/+tSTp66+/Vp8+fbR161ZdfvnlddpfcXGxrFarHA6HQkNDGzN0AGj2yisMDVq8oVoxqopFks0arE/uG8ble/BJns4b2rZtqy+++EI9evTwQHTVVeVNM2fOrLZt+fLlmjdvno4cOdKgY5BLAQCAunA3ZzB1htSpysvLtWrVKh07dkzx8fHasWOHTp48qREjRjj79O7dW126dNHWrVtr3U9ZWZmKi4tdHgAA78jMK6q1GCVJhqR8R6ky89ybGQI0VQMGDNDevXs9vt/T86baHD16VF27dlVkZORZZ1NVIZcCAADeUK9FzT25OOfu3bsVHx+v0tJStWvXTm+//baio6OVnZ2tVq1aqX379i79O3bsqIKCglr3l5qaqpSUFLdiAAB4RmFJ7cWo+vQDmro77rhDv/3tb1VQUKCLLrpILVu2dNnet29ft/ZXW95Uk169eumVV15R37595XA49Ic//EFXXHGFvvzyS3Xu3LnWY5BLAQAAb3D7kr1t27Zp0qRJ+vbbb3X6Sy0Wi8rLy90K4MSJEzpw4IAcDofefPNNvfzyy8rIyFB2drZmzJihsrIyl/5xcXG66qqrtHjx4hr3V1ZW5vKa4uJiRUZGMs0cALxg677DmvjStrP2e33W5YrvHu6FiAD3ePrytICA6pPRLRaLDMPwaN5UW1HqVCdPnlSfPn00ceJEPfLII7X2I5cCAAD14W4e5fYMqarFOd977z3Z7XZZLA1bA6RVq1bOdRX69++v7du3a+nSpbrhhht04sQJHTlyxGWW1KFDh2Sz2WrdX1BQkIKCghoUEwCgfuKiwmS3BqvAUaqafu2oWkMqLirM26EBpsjLy/Po/mrLm1544YWzvrZly5a65JJLznoJIbkUAADwBrcLUt98843efPPNRlucs6KiQmVlZerfv79atmypjz/+WBMmTJAk7dmzRwcOHDjjWgkAAPMEBliUnBit2SuyZJFcilJVP18kJ0azoDmaja5duzbq/qvyprooLy/X7t27NWbMmEaNCQAAoC7cXtTck4tzJiUlafPmzdq/f792796tpKQkbdq0SZMnT5bVatXMmTO1YMECbdy4UTt27NCMGTMUHx9f5zvsAQC8LyHGrrQpsbJZg13abdZgpU2JVUKM3aTIAHO8+uqrGjhwoDp16qRvv/1WkrRkyRKtWbPGrf2cKW+SpKlTpyopKcnZ/+GHH9aHH36o//znP8rKytKUKVP07bff6pZbbvHcmwMAAKgnt2dIeXJxzsLCQk2dOlX5+fmyWq3q27ev1q1bp5EjR0qSnnrqKQUEBGjChAkqKyvTqFGj9Nxzz7kbMgDAyxJi7BoZbVNmXpEKS0oVEVJ5mR4zo9DcpKWl6aGHHtK8efP02GOPOdeMat++vZYsWaKxY8fWeV9ny5sOHDjgsmbVTz/9pFmzZqmgoEAdOnRQ//79tWXLljqtNwUAANDY3F7U3NOLczY2Ty9OCgAA/Jen84bo6Gg9/vjjGjdunEJCQvTFF1/o/PPPV05OjoYOHar//ve/Hoi6cZFLAQCAumj0Rc09vTgnAACAv8rLy9Mll1xSrT0oKEjHjh0zISIAAADf4HZBqrEX5wQAAPAXUVFRys7OrpY/paenq0+fPiZFBQAAYD63FzWXPLc4JwAAgD9bsGCB5syZozfeeEOGYSgzM1OPPfaYkpKSdO+995odHgAAgGncLkilpaVpwYIFGjNmjI4cOVJtcU4AAABUuuWWW7R48WI98MADOn78uCZNmqS0tDQtXbpUN954o9nhAQAAmMbtRc2b2uKcLMQJeFd5hcGd1QA0WY2ZNxw/flxHjx5VRESER/fb2MilAABAXXhlUXMW5wRQk/ScfKWszVW+o9TZZrcGKzkxWgkxdhMjAwDztWnTRm3atDE7DAAAAJ/gdkGKxTkB1CQ9J1+zV2Tp9CmXBY5SzV6RpbQpsRSlADQLl1xyiSyWus0MzcrKauRoAAAAfJPbBamqxTlLS0udi3O+/vrrSk1N1csvv9wYMQLwceUVhlLW5lYrRkmSIckiKWVtrkZG27h8D4DfGzdunPPfpaWleu655xQdHa34+HhJ0rZt2/Tll1/qN7/5jUkRAgAAmM/tgtQtt9yi1q1buyzO2alTJxbnBJqxzLwil8v0TmdIyneUKjOvSPHdw70XGACYIDk52fnvW265RXfeeaceeeSRan0OHjzo7dAAAAB8htsFKUmaPHmyJk+e3GQX5wTgWYUltRej6tMPAPzF6tWr9fnnn1drnzJlii699FK98sorJkQFAABgvoCGvLhNmzYUowAoIiTYo/0AwF+0bt1an376abX2Tz/9VMHB/DcRAAA0X3WaIcXinADOJC4qTHZrsAocpTWuI2WRZLMGKy4qzNuhAYCp5s2bp9mzZysrK0txcXGSpM8++0yvvPKKHnzwQZOjAwAAME+dClIszgngTAIDLEpOjNbsFVmySC5FqapSdnJiNAuaNzHlFYYy84pUWFKqiJDKgiKfIeCe+++/X+eff76WLl2qFStWSJL69OmjZcuW6frrrzc5OgAAAPNYDMOoaUJDrW655RbZ7fZaF+f0tbUQiouLZbVa5XA4FBoaanY4gF9Lz8lXytpclwXO7dZgJSdGKyHGbmJkcBefJZor8obqGBMAAFAX7uYMbhekrFarPv/8c/Xs2dOl/ZtvvtGll14qh8PhXsSNjCQK8C5m1TR96Tn5mr0iq9rll1WfYtqUWIpS8FvkDdUxJgAAoC7czRncvste1eKcpxekWJwTgFR5+V5893Czw0A9lVcYSlmbW+NaYIYqi1Ipa3M1MtpGoRGoRYcOHeq89mZRUVEjRwMAAOCb3C5IsTgnAPivzLwil8v0TmdIyneUKjOviMIjUIslS5Y4/3348GE9+uijGjVqlHPtza1bt2rdunXkTQAAoFlzuyDF4pwA4L8KS2ovRtWnH9AcTZs2zfnvCRMm6OGHH9bcuXOdbXfeeaeeeeYZffTRR5o/f74ZIQIAAJjO7YKUJF1//fUUnwDAD0WE1O3S67r2A5q7devWafHixdXaExISdP/995sQEQAAgG8IMDsAAIDviIsKk90arNpWv7Go8m57cVFh3gwLaLLCw8O1Zs2aau1r1qxReDiXvQIAgOarTjOkWJwTjYm7sgHVmfW9CAywKDkxWrNXZMkiuSxuXnX05MRovqNAHaWkpOiWW27Rpk2bNGDAAEmVa2+mp6frpZdeMjk6AAAA89SpIMXinGgs6Tn5Slmb67KIst0arOTEaG4rj2bL7O9FQoxdaVNiq8Vg47sJuG369Onq06ePnn76ab311luSKtfe/OSTT5wFKgAAgObIYhhGTXf3rtWECRN01VVXuSzOKcm5OOc777zjyfgarLi4WFarVQ6HQ6GhoWaHg1Ok5+Rr9oqsareXr5p3kTYllj980ez40veC2YtojsgbqmNMAABAXbibM7i9htS6deuUkJBQrT0hIUEfffSRu7tDM1VeYShlbW61P7ql/10ilLI2V+UVbtVLgSbN174XgQEWxXcP19h+5ym+ezjFKKCOiouLXf59pgcAAEBz5XZBisU54QmZeUUulwKdzpCU7yhVZh5rkqH54HsB+IcOHTqosLBQktS+fXt16NCh2qOqHQAAoLmq0xpSp2JxTnhCYUntf3TXpx/gD/heAP5hw4YNCgurvBPlxo0bTY4GAADAN7ldkGJxTnhCREiwR/sB/oDvBeAfhgwZUuO/AQAA8D9uF6QkacCAAXrttdc8HQuakbioMNmtwSpwlNa4Xo5FlXf0iosK83ZogGn4XgD+Y9euXXXq17dv30aOBAAAwDfVqSBVXFzsXCH9bAtwunP3ldTUVL311lv6+uuv1bp1a11xxRVavHixevXq5ewzdOhQZWRkuLzutttu0/PPP1/n48D3BAZYlJwYrdkrsmSRXP74rlo2OTkxmkWU0azwvQD8R79+/WSxWFR1M2OLpfJ7e+rNjS0Wi8rLy02JDwAAwGx1Kkh16NBB+fn5ioiIUPv27Z1J1akMw3A7scrIyNCcOXN02WWX6ZdfftHvfvc7XX311crNzVXbtm2d/WbNmqWHH37Y+bxNmzZ1PgZ8V0KMXWlTYpWyNtdlIWebNVjJidFeu7U94Ev4XgD+IS8vz/lvwzAUExOj999/X127djUxKgAAAN9Rp4JUYy3OmZ6e7vJ8+fLlioiI0I4dOzR48GBne5s2bWSz2Tx2XPiOhBi7RkbblJlXpMKSUkWEVF6OxAwQNGd8L4Cm7/TCk8ViUefOnSlIAQAA/H91Kkh5a3FOh8MhSc7iV5XXXntNK1askM1mU2Jioh588EFmSfmRwACL4ruHmx0G4FP4XgAAAADwZ3Ve1LyxF+esqKjQvHnzNHDgQMXExDjbJ02apK5du6pTp07atWuX7rvvPu3Zs8d5h7/TlZWVqayszPn8bGteAQAAAAAAwLvqXJBq7MU558yZo5ycHH3yyScu7bfeeqvz3xdddJHsdruGDx+uffv2qXv37tX2k5qaqpSUlHrFAAAA0FhqWoMTAACguapzQaoxF+ecO3eu3n33XW3evFmdO3c+Y98BAwZIkvbu3VtjQSopKUkLFixwPi8uLlZkZGSDYwQAAKirSy65xKUA9fPPPysxMVGtWrVy6ZeVleXt0AAAAHxCnQtSjbE4p2EYuuOOO/T2229r06ZNioqKOutrsrOzJUl2e813mgoKClJQUFC9YwIAAGiocePGuTwfO3asOYEAAAD4qDoXpBrDnDlztHLlSq1Zs0YhISEqKCiQJFmtVrVu3Vr79u3TypUrNWbMGIWHh2vXrl2aP3++Bg8eXO+1qgAAABpbcnKy2SEAAAD4NFMLUmlpaZKkoUOHurQvW7ZM06dPV6tWrfTRRx9pyZIlOnbsmCIjIzVhwgQ98MADJkQLAAAAAAAAT2hQQaqhi3OeuiB6TSIjI5WRkdGgYwAAAAAAAMC31LkgxeKcAAAAAAAA8IQ6F6RYnBMAAAAAAACeUOeCFItzAgAAmCctLU1paWnav3+/JOnCCy/UQw89pNGjR9f6mtWrV+vBBx/U/v371bNnTy1evFhjxozxUsRoTsorDGXmFamwpFQRIcGKiwpTYEDDlvdA/fBZ+A9/+Cz94T2g8Zi6qDkAAIC/mzt3rh5++GGFhYU1aD+dO3fWokWL1LNnTxmGob/+9a8aO3asdu7cqQsvvLBa/y1btmjixIlKTU3Vtddeq5UrV2rcuHHKyspSTExMg2IBTpWek6+UtbnKd5Q62+zWYCUnRishxm5iZM0Pn4X/8IfP0h/eAxqXxTjbyuJNXHFxsaxWqxwOh0JDQ80OBwDQzPDLYNPiqbzhu+++U+fOnSVJYWFhys7OVpcuXXTRRRfp/fffV2RkpEfiDQsL05NPPqmZM2dW23bDDTfo2LFjevfdd51tl19+ufr166fnn3++zscgl8KZpOfka/aKLJ3+B0XVf+XSpsTyh6eX8Fn4D3/4LP3hPcB97uYMAV6ICQCAZik9J1+DFm/QxJe26a5V2Zr40jYNWrxB6Tn5ZoeGRta7d2917dpVkyZNUmlpqQ4ePChJ2r9/v06ePNng/ZeXl2vVqlU6duyY4uPja+yzdetWjRgxwqVt1KhR2rp1a4OPD0iVBfeUtbnV/uCU5GxLWZur8gq//v3bJ/BZ+A9/+Cz94T3AOyhIAQDQCKp+GTx1mrokFThKNXtFFkUpP3fkyBGtXr1a/fv3V0VFhcaMGaMLLrhAZWVlWrdunQ4dOlSv/e7evVvt2rVTUFCQbr/9dr399tuKjo6usW9BQYE6duzo0taxY0cVFBSc8RhlZWUqLi52eQA1ycwrqvbfuFMZkvIdpcrMK/JeUM0Un4X/8IfP0h/eA7yDghQAAB7GL4M4efKk4uLi9Nvf/latW7fWzp07tWzZMgUGBuqVV15RVFSUevXq5fZ+e/XqpezsbH322WeaPXu2pk2bptzcXI/GnpqaKqvV6nx46vJC+J/Cktr/4KxPP9Qfn4X/8IfP0h/eA7yjXgWpuXPnqqiIaiYAADXhl0G0b99eAwYM0IIFC3TixAn9/PPPGjhwoFq0aKE33nhDP/30k/7yl7+4vd9WrVqpR48e6t+/v1JTU3XxxRdr6dKlNfa12WzVZmIdOnRINpvtjMdISkqSw+FwPqouNwROFxES7NF+qD8+C//hD5+lP7wHeEedC1Lfffed898rV67U0aNHJUkXXXQRiQoAAKfgl0F8//33euCBBxQUFKRffvlF/fv315VXXqkTJ04oKytLFotFgwYNavBxKioqVFZWVuO2+Ph4ffzxxy5t69evr3XNqSpBQUEKDQ11eQA1iYsKk90arNpu02BR5R214qIadodJnB2fhf/wh8/SH94DvKPOBanGXpwTAAB/wS+DOOecc5SYmKjU1FS1adNG27dv1x133CGLxaK7775bVqtVQ4YMcWufSUlJ2rx5s/bv36/du3crKSlJmzZt0uTJkyVJU6dOVVJSkrP/XXfdpfT0dP3xj3/U119/rYULF+rzzz/X3LlzPfpe0XwFBliUnFi5htnpf3hWPU9OjObOol7AZ+E//OGz9If3AO+oc0GqsRbnBADA3/DLIE5ntVp1/fXXq2XLltqwYYPy8vL0m9/8xq19FBYWaurUqerVq5eGDx+u7du3a926dRo5cqQk6cCBA8rP/99i+VdccYVWrlypF198URdffLHefPNNvfPOO4qJifHoe0PzlhBjV9qUWNmsrgV2mzWY27p7GZ+F//CHz9If3gMan8UwjDqtqPrzzz+rdevWkqQOHTpox44dys/P14gRIxQTE6Mvv/xSkZGR2rNnT6MG7K7i4mJZrVY5HA6mnAMAvKbqLnuSXBY3rypSkYz5psbIGw4ePKjzzjtPAQEBiomJ0QcffNCkFgonl0JdlFcYyswrUmFJqSJCKgvuzH4wB5+F//CHz9If3gPqzt2coUVdd9y+fXv169dPAwcOrHFxzvPOO0/bt29vUPAAAPiLql8GU9bmuixwbrMGKzkxmmJUM3Jq8SknJ8fESIDGExhgUXz3cLPDgPgs/Ik/fJb+8B7QeOpckPr++++1detWbdmyxbk452WXXeZcnLNz584eWZwTAAB/kRBj18hoG78MAgAAAKep8yV7p+rQoYM2b96sr776SlOnTnXeVjguLk4ZGRmNEWe9Mc0cAADUFXlDdYwJAACoC3dzhjovan46TyzOCQAAAAAAgOanzpfsnWrXrl0677zzJEldu3ZVy5YtZbPZdMMNN3g0OAAAAAAAAPifehWkWJwTAAAAAAAA9VXvS/YAAAAAAACA+qAgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAArzK1IJWamqrLLrtMISEhioiI0Lhx47Rnzx6XPqWlpZozZ47Cw8PVrl07TZgwQYcOHTIpYgAAAAAAADSUqQWpjIwMzZkzR9u2bdP69et18uRJXX311Tp27Jizz/z587V27VqtXr1aGRkZ+uGHHzR+/HgTowYAAAAAAEBDWAzDMMwOosqPP/6oiIgIZWRkaPDgwXI4HDr33HO1cuVK/frXv5Ykff311+rTp4+2bt2qyy+//Kz7LC4ultVqlcPhUGhoaGO/BQAA0ISRN1THmAAAgLpwN2fwqTWkHA6HJCksLEyStGPHDp08eVIjRoxw9undu7e6dOmirVu31riPsrIyFRcXuzwAAAAAAADgO3ymIFVRUaF58+Zp4MCBiomJkSQVFBSoVatWat++vUvfjh07qqCgoMb9pKamymq1Oh+RkZGNHToAAAAAAADc4DMFqTlz5ignJ0erVq1q0H6SkpLkcDicj4MHD3ooQgAAAAAAAHhCC7MDkKS5c+fq3Xff1ebNm9W5c2dnu81m04kTJ3TkyBGXWVKHDh2SzWarcV9BQUEKCgpq7JABAAAAAABQT6bOkDIMQ3PnztXbb7+tDRs2KCoqymV7//791bJlS3388cfOtj179ujAgQOKj4/3drgAvKC8wtDWfYe1Jvt7bd13WOUVPnPfBcA0fC8AAADgb0ydITVnzhytXLlSa9asUUhIiHNdKKvVqtatW8tqtWrmzJlasGCBwsLCFBoaqjvuuEPx8fF1usMegKYlPSdfKWtzle8odbbZrcFKToxWQozdxMgA8/C9AAAAgD+yGIZh2s+sFoulxvZly5Zp+vTpkqTS0lL99re/1euvv66ysjKNGjVKzz33XK2X7J2OWxUDTUN6Tr5mr8jS6f9BqvqvRNqUWP74RrPD98L7yBuqY0wAAEBduJszmFqQ8gaSKMD3lVcYGrR4g8sMkFNZJNmswfrkvmEKDKi5kA34G74X5iBvqI4xAQAAdeFuzuAzd9kD0Hxl5hXV+ke3JBmS8h2lyswr8l5QgMn4XgAAAMCfUZACYLrCktr/6K5PP8Af8L0AAACAP6MgBcB0ESHBHu0H+AO+FwAAAPBnFKQAD+G27PUXFxUmuzVYta2CY1HlXcXiosK8GRZgKr4XAAAA8GctzA4A8Afclr1hAgMsSk6M1uwVWbJILncUq/pjPDkxmoWb0azwvQAAAIA/Y4YU0EBVt2U/ffHhAkepZq/IUnpOvkmRNS0JMXalTYmVzep6+ZHNGsyt7dFs8b0AAACAv2KGFNAA5RWGUtbmqqaL8wxVzmJIWZurkdE2ZjHUQUKMXSOjbcrMK1JhSakiQiovR2Ls0JzxvQAAAIA/oiAFNIA7t2WP7x7uvcCasMAAC2MFnIbvBQAAAPwNl+wBDcBt2QEAAAAAcB8FKaABuC07AAAAAADuoyAFNAC3ZQcAAAAAwH0UpIAGqLotu6RqRSluyw7An5RXGNq677DWZH+vrfsOq7yipts5AAAAAHXDouZAA1Xdlj1lba7LAuc2a7CSE6O5LTuAJi89J7/af+Ps/DcOAAAADUBBCvAAbssOwF+l5+Rr9oosnT4fqsBRqtkrspQ2JZaiFAAAANxGQQrwEG7LDsDflFcYSlmbW60YJUmGKi9NTlmbq5HRNgrwAAAAcAtrSAEAgBpl5hW5XKZ3OkNSvqNUmXlF3guqGUtNTdVll12mkJAQRUREaNy4cdqzZ88ZX7N8+XJZLBaXR3Awd34FAADmoyAFAABqVFhSezGqPv3QMBkZGZozZ462bdum9evX6+TJk7r66qt17NixM74uNDRU+fn5zse3337rpYgBAABqxyV7AACgRhEhdZtJU9d+aJj09HSX58uXL1dERIR27NihwYMH1/o6i8Uim83W2OEBAAC4hRlSAACgRnFRYbJbg1Xb6lAWVd5tLy4qzJth4f9zOBySpLCwM4//0aNH1bVrV0VGRmrs2LH68ssvz9i/rKxMxcXFLg8AAABPoyAFAICfK68wtHXfYa3J/l5b9x1WeUVNy5RXFxhgUXJitCRVK0pVPU9OjGZBcxNUVFRo3rx5GjhwoGJiYmrt16tXL73yyitas2aNVqxYoYqKCl1xxRX67rvvan1NamqqrFar8xEZGdkYbwEAADRzFsMw6paVNlHFxcWyWq1yOBwKDQ01OxwAALwqPSdfKWtzXRYnt1uDlZwYrYQYu9f20VQ0lbxh9uzZ+uCDD/TJJ5+oc+fOdX7dyZMn1adPH02cOFGPPPJIjX3KyspUVlbmfF5cXKzIyEifHxMAAGAud/Mo1pACAMBPpefka/aKLJ3+y1OBo1SzV2QpbUpsnQpKCTF2jYy2KTOvSIUlpYoIqbxMj5lR5pg7d67effddbd682a1ilCS1bNlSl1xyifbu3Vtrn6CgIAUFBTU0TAAAgDPikj0AAPxQeYWhlLW51YpRkpxtKWtz3bp8L757uMb2O0/x3cMpRpnAMAzNnTtXb7/9tjZs2KCoqCi391FeXq7du3fLbvevmW0AAKDpoSAFAIAfyswrcrnE7nSGpHxHqTLzirwXFBpkzpw5WrFihVauXKmQkBAVFBSooKBAP//8s7PP1KlTlZSU5Hz+8MMP68MPP9R//vMfZWVlacqUKfr22291yy23mPEWAAAAnLhkDwAAP1RYUnsxqj79YL60tDRJ0tChQ13aly1bpunTp0uSDhw4oICA//3e+NNPP2nWrFkqKChQhw4d1L9/f23ZskXR0dHeChsAAKBGFKQAAPBDESHBHu0H89XlPjSbNm1yef7UU0/pqaeeaqSIAAAA6o9L9gAA8ENxUWGyW4NV20pPFlXeKS8uKsybYQEAAACSKEgBAHxYeYWhrfsOa03299q673CdF+BG5SLkyYmVl2WdXpSqep6cGM3i5AAAADCFqQWpzZs3KzExUZ06dZLFYtE777zjsn369OmyWCwuj4SEBHOCBQB4VXpOvgYt3qCJL23TXauyNfGlbRq0eIPSc/LNDq3JSIixK21KrGxW18vybNZgpU2JVUIMd1oDAACAOUxdQ+rYsWO6+OKLdfPNN2v8+PE19klISNCyZcucz4OCgrwVHgDAJOk5+Zq9Ikunz4cqcJRq9oosiiluSIixa2S0TZl5RSosKVVESOVlesyMAgAAgJlMLUiNHj1ao0ePPmOfoKAg2Ww2L0UEADBbeYWhlLW51YpRkmSo8nKzlLW5Ghlto6hSR4EBFsV3Dzc7DAAAAMDJ59eQ2rRpkyIiItSrVy/Nnj1bhw8fPmP/srIyFRcXuzwAAE1HZl6R8h2ltW43JOU7SpWZV+S9oAAAAAB4lE8XpBISEvS3v/1NH3/8sRYvXqyMjAyNHj1a5eXltb4mNTVVVqvV+YiMjPRixACAhiosqb0YVZ9+AAAAAHyPqZfsnc2NN97o/PdFF12kvn37qnv37tq0aZOGDx9e42uSkpK0YMEC5/Pi4mKKUgDQhESEBJ+9kxv9AAAAAPgen54hdbrzzz9f55xzjvbu3Vtrn6CgIIWGhro8AABNR1xUmOzWYNW2OpRFkt1auTA3AAAAgKapSRWkvvvuOx0+fFh2O3dWAuC7yisMbd13WGuyv9fWfYdVXlHT8tyoTWCARcmJ0ZJUrShV9Tw5MZoFzQEAAIAmzNRL9o4ePeoy2ykvL0/Z2dkKCwtTWFiYUlJSNGHCBNlsNu3bt0/33nuvevTooVGjRpkYNQDULj0nXylrc10W5bZbg5WcGK2EGIrpdZUQY1falNhqY2ljLAEAAAC/YDEMw7Sf7jdt2qSrrrqqWvu0adOUlpamcePGaefOnTpy5Ig6deqkq6++Wo888og6duxY52MUFxfLarXK4XBw+R6ARpWek6/ZK7J0+n9Uq+bxpE2JpZDipvIKQ5l5RSosKVVESOVlesyMQmMib6iOMQEAAHXhbs5g6gypoUOH6kz1sHXr1nkxGgCov/IKQylrc6sVoyTJUGVRKmVtrkZG2yiouCEwwKL47uFmhwEAAADAw5rUGlIA4Ksy84pcLi07nSEp31GqzLwi7wUFAAAAAD6KghQAeEBhSe3FqPr0AwAAAAB/RkEKADwgIiTYo/0AAAAAwJ9RkAIAD4iLCpPdGqzaVoeyqPJue3FRYd4MCwAAAAB8EgUpAPCAwACLkhOjJalaUarqeXJiNAuaAwAAAIAoSAGAxyTE2JU2JVY2q+tleTZrsNKmxCohxm5SZAAAAADgW1qYHQAA+JOEGLtGRtuUmVekwpJSRYRUXqbHzCgAAAAA+B8KUgDgYYEBFsV3Dzc7DAAAAADwWVyyBwAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvamF2AIAnlFcYyswrUmFJqSJCghUXFabAAIvZYQEAAAAAgBpQkEKTl56Tr5S1ucp3lDrb7NZgJSdGKyHGbmJkAAAAAACgJlyyhyYtPSdfs1dkuRSjJKnAUarZK7KUnpNvUmQAAAAAAKA2FKTQZJVXGEpZmyujhm1VbSlrc1VeUVMPAAAAAABgFgpSaLIy84qqzYw6lSEp31GqzLwi7wUFAAAAAADOioIUmqzCktqLUfXpBwAAAAAAvIOCFJqsiJBgj/YDAAAAAADeQUEKTVZcVJjs1mBZatluUeXd9uKiwrwZFgAAAAAAOAsKUs1ceYWhrfsOa03299q673CTWgA8MMCi5MRoSapWlKp6npwYrcCA2kpWAACgKWtoHtOU8yAAAJo6UwtSmzdvVmJiojp16iSLxaJ33nnHZbthGHrooYdkt9vVunVrjRgxQt988405wfqh9Jx8DVq8QRNf2qa7VmVr4kvbNGjxBqXn5JsdWp0lxNiVNiVWNqvrZXk2a7DSpsQqIcZuUmQAAHhWamqqLrvsMoWEhCgiIkLjxo3Tnj17zvq61atXq3fv3goODtZFF12k999/3wvRNr6G5jH+kAcBANCUmVqQOnbsmC6++GI9++yzNW5/4okn9PTTT+v555/XZ599prZt22rUqFEqLWWR6oZKz8nX7BVZ1e5SV+Ao1ewVWU0qGUuIseuT+4bp9VmXa+mN/fT6rMv1yX3DKEYBAPxKRkaG5syZo23btmn9+vU6efKkrr76ah07dqzW12zZskUTJ07UzJkztXPnTo0bN07jxo1TTk6OFyP3vIbmMf6UBwEA0FRZDMPwibnJFotFb7/9tsaNGyepcnZUp06d9Nvf/lZ33323JMnhcKhjx45avny5brzxxjrtt7i4WFarVQ6HQ6GhoY0VfpNSXmFo0OIN1ZKwKhZVzjD65L5hXO4GAGhWmlLe8OOPPyoiIkIZGRkaPHhwjX1uuOEGHTt2TO+++66z7fLLL1e/fv30/PPP1+k4vjYmDc1jyIMAAGgc7uYMPruGVF5engoKCjRixAhnm9Vq1YABA7R169ZaX1dWVqbi4mKXB1xl5hXVmoRJkiEp31GqzLwi7wUFAADc4nA4JElhYbXfvGPr1q0uuZQkjRo1qknnUg3NY8iDAADwDT5bkCooKJAkdezY0aW9Y8eOzm01SU1NldVqdT4iIyMbNc6mqLCkbpc81rUfAADwroqKCs2bN08DBw5UTExMrf0KCgr8LpdqaB5DHgQAgG/w2YJUfSUlJcnhcDgfBw8eNDsknxMREnz2Tm70AwAA3jVnzhzl5ORo1apVHt+3r+dSDc1jyIMAAPANLcwOoDY2m02SdOjQIdnt/1uc+tChQ+rXr1+trwsKClJQUFBjh9ekxUWFyW4NVoGjVDUtIFa1dkJcVO2XAAAAAHPMnTtX7777rjZv3qzOnTufsa/NZtOhQ4dc2g4dOuTMs2ri67lUQ/MY8iAAAHyDz86QioqKks1m08cff+xsKy4u1meffab4+HgTI2v6AgMsSk6MllSZdJ2q6nlyYjQLeQIA4EMMw9DcuXP19ttva8OGDYqKijrra+Lj411yKUlav359k86lGprHkAcBAOAbTC1IHT16VNnZ2crOzpZUuZB5dna2Dhw4IIvFonnz5unRRx/VP//5T+3evVtTp05Vp06dnHfiQ/0lxNiVNiVWNqvrdHSbNVhpU2KVEGOv5ZUAAMAMc+bM0YoVK7Ry5UqFhISooKBABQUF+vnnn519pk6dqqSkJOfzu+66S+np6frjH/+or7/+WgsXLtTnn3+uuXPnmvEWPKaheQx5EAAA5rMYhlHTbGWv2LRpk6666qpq7dOmTdPy5ctlGIaSk5P14osv6siRIxo0aJCee+45XXDBBXU+hq/dqtjXlFcYyswrUmFJqSJCKqen84sgAKC58uW8wWKp+f/Py5Yt0/Tp0yVJQ4cOVbdu3bR8+XLn9tWrV+uBBx7Q/v371bNnTz3xxBMaM2ZMnY/ry2PS0DyGPAgAAM9xN2cwtSDlDb6cRAEAAN9C3lAdYwIAAOrC3ZzBZ9eQAgAAAAAAgH+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvamF2AI3NMAxJUnFxscmRAAAAX1eVL1TlDyCXAgAAdeNuHuX3BamSkhJJUmRkpMmRAACApqKkpERWq9XsMHwCuRQAAHBHXfMoi+HnPwFWVFTohx9+UEhIiCwWi8f3X1xcrMjISB08eFChoaEe339zwTh6DmPpGYyjZzCOnsNYesbZxtEwDJWUlKhTp04KCGBlA6lxcynOa89hLD2DcfQcxtIzGEfPYBw950xj6W4e5fczpAICAtS5c+dGP05oaCgntgcwjp7DWHoG4+gZjKPnMJaecaZxZGaUK2/kUpzXnsNYegbj6DmMpWcwjp7BOHpObWPpTh7FT38AAAAAAADwKgpSAAAAAAAA8CoKUg0UFBSk5ORkBQUFmR1Kk8Y4eg5j6RmMo2cwjp7DWHoG4+hb+Dw8h7H0DMbRcxhLz2AcPYNx9BxPjqXfL2oOAAAAAAAA38IMKQAAAAAAAHgVBSkAAAAAAAB4FQUpAAAAAAAAeBUFKQAAAAAAAHgVBakGePbZZ9WtWzcFBwdrwIAByszMNDukJmfhwoWyWCwuj969e5sdls/bvHmzEhMT1alTJ1ksFr3zzjsu2w3D0EMPPSS73a7WrVtrxIgR+uabb8wJ1sedbSynT59e7RxNSEgwJ1gflpqaqssuu0whISGKiIjQuHHjtGfPHpc+paWlmjNnjsLDw9WuXTtNmDBBhw4dMili31SXcRw6dGi1c/L22283KWLflJaWpr59+yo0NFShoaGKj4/XBx984NzOueg7yKUajlyqfsilPIM8yjPIozyHXMozvJVLUZCqpzfeeEMLFixQcnKysrKydPHFF2vUqFEqLCw0O7Qm58ILL1R+fr7z8cknn5gdks87duyYLr74Yj377LM1bn/iiSf09NNP6/nnn9dnn32mtm3batSoUSotLfVypL7vbGMpSQkJCS7n6Ouvv+7FCJuGjIwMzZkzR9u2bdP69et18uRJXX311Tp27Jizz/z587V27VqtXr1aGRkZ+uGHHzR+/HgTo/Y9dRlHSZo1a5bLOfnEE0+YFLFv6ty5sxYtWqQdO3bo888/17BhwzR27Fh9+eWXkjgXfQW5lOeQS7mPXMozyKM8gzzKc8ilPMNruZSBeomLizPmzJnjfF5eXm506tTJSE1NNTGqpic5Odm4+OKLzQ6jSZNkvP32287nFRUVhs1mM5588kln25EjR4ygoCDj9ddfNyHCpuP0sTQMw5g2bZoxduxYU+JpygoLCw1JRkZGhmEYledgy5YtjdWrVzv7fPXVV4YkY+vWrWaF6fNOH0fDMIwhQ4YYd911l3lBNVEdOnQwXn75Zc5FH0Iu5RnkUg1HLuUZ5FGeQx7lOeRSntMYuRQzpOrhxIkT2rFjh0aMGOFsCwgI0IgRI7R161YTI2uavvnmG3Xq1Ennn3++Jk+erAMHDpgdUpOWl5engoICl/PTarVqwIABnJ/1tGnTJkVERKhXr16aPXu2Dh8+bHZIPs/hcEiSwsLCJEk7duzQyZMnXc7L3r17q0uXLpyXZ3D6OFZ57bXXdM455ygmJkZJSUk6fvy4GeE1CeXl5Vq1apWOHTum+Ph4zkUfQS7lWeRSnkUu5VnkUe4jj/IccqmGa8xcqoWng20O/vvf/6q8vFwdO3Z0ae/YsaO+/vprk6JqmgYMGKDly5erV69eys/PV0pKiq688krl5OQoJCTE7PCapIKCAkmq8fys2oa6S0hI0Pjx4xUVFaV9+/bpd7/7nUaPHq2tW7cqMDDQ7PB8UkVFhebNm6eBAwcqJiZGUuV52apVK7Vv396lL+dl7WoaR0maNGmSunbtqk6dOmnXrl267777tGfPHr311lsmRut7du/erfj4eJWWlqpdu3Z6++23FR0drezsbM5FH0Au5TnkUp5HLuU55FHuI4/yHHKphvFGLkVBCqYaPXq08999+/bVgAED1LVrV/3973/XzJkzTYwMqHTjjTc6/33RRRepb9++6t69uzZt2qThw4ebGJnvmjNnjnJycljDpIFqG8dbb73V+e+LLrpIdrtdw4cP1759+9S9e3dvh+mzevXqpezsbDkcDr355puaNm2aMjIyzA4L8DhyKfgy8ij3kUd5DrlUw3gjl+KSvXo455xzFBgYWG0V+UOHDslms5kUlX9o3769LrjgAu3du9fsUJqsqnOQ87NxnH/++TrnnHM4R2sxd+5cvfvuu9q4caM6d+7sbLfZbDpx4oSOHDni0p/zsma1jWNNBgwYIEmck6dp1aqVevToof79+ys1NVUXX3yxli5dyrnoI8ilGg+5VMORSzUe8qgzI4/yHHKphvNGLkVBqh5atWql/v376+OPP3a2VVRU6OOPP1Z8fLyJkTV9R48e1b59+2S3280OpcmKioqSzWZzOT+Li4v12WefcX56wHfffafDhw9zjp7GMAzNnTtXb7/9tjZs2KCoqCiX7f3791fLli1dzss9e/bowIEDnJenONs41iQ7O1uSOCfPoqKiQmVlZZyLPoJcqvGQSzUcuVTjIY+qGXmU55BLNZ7GyKW4ZK+eFixYoGnTpunSSy9VXFyclixZomPHjmnGjBlmh9ak3H333UpMTFTXrl31ww8/KDk5WYGBgZo4caLZofm0o0ePulTw8/LylJ2drbCwMHXp0kXz5s3To48+qp49eyoqKkoPPvigOnXqpHHjxpkXtI8601iGhYUpJSVFEyZMkM1m0759+3TvvfeqR48eGjVqlIlR+545c+Zo5cqVWrNmjUJCQpzXj1utVrVu3VpWq1UzZ87UggULFBYWptDQUN1xxx2Kj4/X5ZdfbnL0vuNs47hv3z6tXLlSY8aMUXh4uHbt2qX58+dr8ODB6tu3r8nR+46kpCSNHj1aXbp0UUlJiVauXKlNmzZp3bp1nIs+hFzKM8il6odcyjPIozyDPMpzyKU8w2u5lCdvA9jc/PnPfza6dOlitGrVyoiLizO2bdtmdkhNzg033GDY7XajVatWxnnnnWfccMMNxt69e80Oy+dt3LjRkFTtMW3aNMMwKm9X/OCDDxodO3Y0goKCjOHDhxt79uwxN2gfdaaxPH78uHH11Vcb5557rtGyZUuja9euxqxZs4yCggKzw/Y5NY2hJGPZsmXOPj///LPxm9/8xujQoYPRpk0b41e/+pWRn59vXtA+6GzjeODAAWPw4MFGWFiYERQUZPTo0cO45557DIfDYW7gPubmm282unbtarRq1co499xzjeHDhxsffvihczvnou8gl2o4cqn6IZfyDPIozyCP8hxyKc/wVi5lMQzDcK+EBQAAAAAAANQfa0gBAAAAAADAqyhIAQAAAAAAwKsoSAEAAAAAAMCrKEgBAAAAAADAqyhIAQAAAAAAwKsoSAEAAAAAAMCrKEgBAAAAAADAqyhIAWhWNm3aJIvFoiNHjtT5Nd26ddOSJUsaLSYAAICmgDwKgCdRkALgM6ZPny6LxaLbb7+92rY5c+bIYrFo+vTp3g/MDRUVFQoNDdW///1vSdIFF1ygzZs3mxwVAADwd+RRAJoaClIAfEpkZKRWrVqln3/+2dlWWlqqlStXqkuXLiZGVjc5OTkKDg7WBRdcoEOHDunbb7/VZZddZnZYAACgGSCPAtCUUJAC4FNiY2MVGRmpt956y9n21ltvqUuXLrrkkktc+paVlenOO+9URESEgoODNWjQIG3fvt2lz/vvv68LLrhArVu31lVXXaX9+/dXO+Ynn3yiK6+8Uq1bt1ZkZKTuvPNOHTt2rF7xb9myRVdccYVzv5dccolat25dr30BAAC4gzwKQFNCQQqAz7n55pu1bNky5/NXXnlFM2bMqNbv3nvv1T/+8Q/99a9/VVZWlnr06KFRo0apqKhIknTw4EGNHz9eiYmJys7O1i233KL777/fZR/79u1TQkKCJkyYoF27dumNN97QJ598orlz57oVc/v27dW+fXvNmzdPH3zwgdq3b68pU6YoKytL7du317XXXluPkQAAAHAPeRSApoKCFACfM2XKFH3yySf69ttv9e233+rTTz/VlClTXPocO3ZMaWlpevLJJzV69GhFR0frpZdeUuvWrfWXv/xFkpSWlqbu3bvrj3/8o3r16qXJkydXWzshNTVVkydP1rx589SzZ09dccUVevrpp/W3v/1NpaWldY45OztbO3fuVJs2bbRmzRplZ2crMjJSy5cvV3Z2tl5++eUGjwsAAMDZkEcBaCpamB0AAJzu3HPP1TXXXKPly5fLMAxdc801Ouecc1z67Nu3TydPntTAgQOdbS1btlRcXJy++uorSdJXX32lAQMGuLwuPj7e5fkXX3yhXbt26bXXXnO2GYahiooK5eXlqU+fPnWKuVu3bsrMzFSbNm2UkJCg7777Tj/88IMmTJigoKAgt94/AABAfZFHAWgqKEgB8Ek333yzc7r3s88+22jHOXr0qG677Tbdeeed1bbVdfHP0aNH61//+pd++eUX/fLLL2rXrp3Ky8tVVlam8PBw53EAAAC8gTwKQFNAQQqAT0pISNCJEydksVg0atSoatu7d++uVq1a6dNPP1XXrl0lSSdPntT27ds1b948SVKfPn30z3/+0+V127Ztc3keGxur3Nxc9ejRo96xvvzyy/r55581bdo0jR8/XmPHjtXdd9+t3r1765Zbbqn3fgEAAOqDPApAU8AaUgB8UmBgoL766ivl5uYqMDCw2va2bdtq9uzZuueee5Senq7c3FzNmjVLx48f18yZMyVJt99+u7755hvdc8892rNnj1auXKnly5e77Oe+++7Tli1bNHfuXGVnZ+ubb77RmjVr3FqM87zzzlO3bt20a9cujR8/Xj169NCuXbs0duxY9ejRo0FJGgAAgLvIowA0BRSkAPis0NBQhYaG1rp90aJFmjBhgm666SbFxsZq7969WrdunTp06CCpcqr4P/7xD73zzju6+OKL9fzzz+vxxx932Uffvn2VkZGhf//737ryyit1ySWX6KGHHlKnTp3civXzzz9X+/btFRUVpe+++06HDh3SpZde6v6bBgAA8ADyKAC+zmIYhmF2EAAAAAAAAGg+mCEFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACvoiAFAAAAAAAAr6IgBQAAAAAAAK+iIAUAAAAAAACv+n84MSkW83corgAAAABJRU5ErkJggg==\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZT2iKkmKUkd6"
      },
      "source": [
        "## MLR"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1mN-6wdWUdwN"
      },
      "outputs": [],
      "source": [
        "\n",
        "# coding: utf-8\n",
        "#import all required libraries etc\n",
        "from scipy.io import loadmat\n",
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "from sklearn import linear_model\n",
        "import statsmodels.api as sm\n",
        "import itertools\n",
        "# In[3]:\n",
        "#run notebook with functions we'll need\n",
        "get_ipython().run_line_magic('run', 'Lab2_2023_functions.ipynb')\n",
        "# In[4]:\n",
        "#load the data and check it out\n",
        "data = pd.read_csv('glaciers.csv')\n",
        "data\n",
        "# In[5]:\n",
        "### First, we need to clean the data -- g<=0 is non physical and should be removed.\n",
        "### Are there any points rows that meet this condition?\n",
        "np.where(data[\"g\"]<=0)\n",
        "# In[6]:\n",
        "### Yes, one row should be removed.\n",
        "data = data[data[\"g\"]>0].reset_index(drop=True)\n",
        "### data[\"g\"]>0 generates a series containing \"True\" and \"False\" used to index data\n",
        "### and remove \"False\" entries\n",
        "# In[7]:\n",
        "data # Notice that it is one row shorter!\n",
        "# In[8]:\n",
        "#normalize data and check it out\n",
        "data_norm = (data - data.mean())/data.std()\n",
        "data_norm\n",
        "# In[9]:\n",
        "#visualize normalization\n",
        "plt.figure(figsize=[15,4])\n",
        "plt.subplot(1,2,1)\n",
        "plt.plot(data['g'])\n",
        "plt.plot(data_norm['g'])\n",
        "plt.legend(['Raw data','Normalized'])\n",
        "plt.ylabel('g')\n",
        "plt.xlabel('glacier')\n",
        "plt.title('Visualizing Normalization')\n",
        "plt.subplot(1,2,2)\n",
        "plt.scatter(data['g'],data_norm['g'])\n",
        "plt.ylabel('g_norm')\n",
        "plt.xlabel('g')\n",
        "plt.title('Visualizing Normalization')\n",
        "# In[10]:\n",
        "#unpack normalized data\n",
        "lat = data_norm['lat']\n",
        "lon = data_norm['lon']\n",
        "hmax = data_norm['max_elevation']\n",
        "hmed = data_norm['median_elevation']\n",
        "g = data['g'] # response variable does not need to be standardized\n",
        "Pannual = data_norm['precipitation']\n",
        "Pwinter = data_norm['winter_precipitation']\n",
        "CI = data_norm['continentality']\n",
        "cloud = data_norm['cloud_cover']\n",
        "Tsummer = data_norm['summer_temperature']\n",
        "# In[11]:\n",
        "#target variable: y; predictor variable(s): x\n",
        "y = g\n",
        "X = data_norm.drop('g',axis=1)\n",
        "# # SOLUTIONS:\n",
        "# # 1) Apply MLR to standardized input\n",
        "# In[12]:\n",
        "#make MLR model, fit to the data, and predict y\n",
        "lm_MLR = linear_model.LinearRegression()\n",
        "model = lm_MLR.fit(X,y)\n",
        "ypred_MLR = lm_MLR.predict(X) #y predicted by MLR\n",
        "intercept_MLR = lm_MLR.intercept_ #intercept predicted by MLR\n",
        "coef_MLR = lm_MLR.coef_ #regression coefficients in MLR model\n",
        "R2_MLR = lm_MLR.score(X,y) #R-squared value from MLR model\n",
        "corr_coef = np.corrcoef(ypred_MLR,y) # Pearson correlation between modeled andtarget y\n",
        "print('correlation = ' + str(corr_coef[0, 1]))\n",
        "# In[13]:\n",
        "#visualize MLR model performance\n",
        "ax = plt.subplot()\n",
        "ax.scatter(y,ypred_MLR)\n",
        "### We want to draw a 1:1 line but don't know the lims to draw it over -- so we query the plot lims,\n",
        "### Draw a line accordingly and ensure to set the lims\n",
        "lims = ax.get_xlim()\n",
        "ax.set_xlim(lims)\n",
        "ax.set_ylim(lims)\n",
        "ax.plot(lims,lims,ls=\"--\", c=\"grey\")\n",
        "ax.set_xlabel('Measured g')\n",
        "ax.set_ylabel('Modelled g')\n",
        "ax.set_title(f'MLR Model Results: R$^2$ = {round(R2_MLR,2)}')\n",
        "#do MLR using predictors chosen from stepwise regression\n",
        "lm_step = linear_model.LinearRegression()\n",
        "model_step = lm_step.fit(X[result],y)\n",
        "ypred_step = lm_step.predict(X[result]) #y predicted by MLR\n",
        "intercept_step = lm_step.intercept_ #intercept predicted by MLR\n",
        "coef_step = lm_step.coef_ #regression coefficients in MLR model\n",
        "R2_step = lm_step.score(X[result],y) #R-squared value from MLR model\n",
        "# In[17]:\n",
        "#visualize stepwise model performance\n",
        "ax = plt.subplot()\n",
        "ax.scatter(y,ypred_step)\n",
        "lims = ax.get_xlim()\n",
        "ax.set_xlim(lims)\n",
        "ax.set_ylim(lims)\n",
        "ax.plot(lims,lims,ls=\"--\", c=\"grey\")\n",
        "ax.set_xlabel('Measured g')\n",
        "ax.set_ylabel('Modelled g')\n",
        "ax.set_title(f'Stepwise Model Results: R$^2$ = {round(R2_step,2)}')\n",
        "# Apply calibration - validation approach on standardized input\n",
        "# In[18]:\n",
        "#goal: loop through every combination of normalized predictors, make linear model, and find one with best performance\n",
        "R2_best = []\n",
        "combo_best = []\n",
        "for kk in range(1,10): #for each total number of predictors to use in model (from 1predictor to 9 predictors)\n",
        "v0 = range(np.shape(X)[1])\n",
        "combinations = list(itertools.combinations(range(np.shape(X)[1]),kk)) #allpossible combinations of kk total predictors\n",
        "R2_test = []\n",
        "for ind in range(len(combinations)): #for each combination of predictors, makeMLR model and compute R^2\n",
        "test_vars = np.array(combinations[ind])\n",
        "X_test = X.iloc[::2,test_vars] # calibation sample consists of all oddindices in the data\n",
        "y_test = y.iloc[::2]\n",
        "X_valid = X.iloc[1::2,test_vars] # validation sample consists of all evenindices in the data\n",
        "y_valid = y.iloc[1::2]\n",
        "lm_test = linear_model.LinearRegression()\n",
        "model_test = lm_test.fit(X_test,y_test)\n",
        "ypred_test = lm_test.predict(X_test) #y predicted by MLR\n",
        "R2_test.append(lm_test.score(X_valid,y_valid)) #R-squared value from MLR\n",
        "model\n",
        "R2_best.append(np.max(R2_test))\n",
        "combo_best.append(combinations[np.argmax(R2_test)])\n",
        "R2_best_final = np.max(R2_best)\n",
        "combo_best_final = combo_best[np.argmax(R2_best)]\n",
        "print('The best combination of predictors is: ' )\n",
        "print(list(X.columns[np.asarray(combo_best_final)]))\n",
        "# In[19]:\n",
        "#build linear model using the best combination of predictors\n",
        "X_calib_valid = X.iloc[:,np.asarray(combo_best_final)]\n",
        "lm_calib_valid = linear_model.LinearRegression()\n",
        "model_calib_valid = lm_calib_valid.fit(X_calib_valid,y)\n",
        "ypred_calib_valid = lm_calib_valid.predict(X_calib_valid) #y predicted by MLR\n",
        "intercept_calib_valid = lm_calib_valid.intercept_ #intercept predicted by MLR\n",
        "coef_calib_valid = lm_calib_valid.coef_ #regression coefficients in MLR model\n",
        "R2_calib_valid = lm_calib_valid.score(X_calib_valid,y) #R-squared value from MLR\n",
        "model\n",
        "# In[20]:\n",
        "#visualize calibration-validation model performance\n",
        "ax = plt.subplot()\n",
        "ax.scatter(y,ypred_calib_valid)\n",
        "lims = ax.get_xlim()\n",
        "ax.set_xlim(lims)\n",
        "ax.set_ylim(lims)\n",
        "ax.plot(lims,lims,ls=\"--\", c=\"grey\")\n",
        "ax.set_xlabel('Measured g')\n",
        "ax.set_ylabel('Modelled g')\n",
        "ax.set_title(f'Calibration-Validation Model Results: R$^2$ ={round(R2_calib_valid,2)}')\n",
        "# In[21]:\n",
        "#Answers to the specific questions:\n",
        "# 1) How well can the spatial variability in mass balance gradient be represented\n",
        "# by a multiple linear regression (MLR) using all the plausible predictors from thegiven dataset?\n",
        "# Answer: R^2 is a metric of the goodness of the fit, which is the same ascorrelation\n",
        "# coefficient squared. In this case, R^2 = 0.38 (38% variance is expained by theMLR model)\n",
        "#2) What is the optimal number of predictors in MLR and how much variance in the mass balance gradient\n",
        "# can be explained by this MLR?\n",
        "# Answers: According to the stepwise regression there is only one optimal predictor: CI (continentality index)\n",
        "# With this one predictor, the model explains 25% of the variance (R^2 = 0.25)\n",
        "# According to the calibration-validation approach optimal number of preditors is4: latitude, max elevation, median elevation, cloud cover\n",
        "# With this four-predictor model R^2 = 0.32 (32% variance is explained)\n",
        "# 3) Plot modelled (regressed) mass balance gradients versus observed ones on ascatter plot (with 1:1 line)\n",
        "# and discuss whether the linear model represents a good fit to the data or not.\n",
        "# The plots are provided and no obvius outliers are indicated in the scatter plot\n",
        "# There seems to be a tendency for the model (in all three cases above) to\n",
        "# underestimate mass balance gradients that are larger then 10 mm w.e. per m\n",
        "# # Saving figures in python:\n",
        "# To save figures, use the function plt.savefig()\n",
        "#\n",
        "# Vector graphics (.pdf, .eps, etc.) are generally preferable to raster graphics(.jpg, .png) because they are not defined by pixels but instead are defined with aset of \"drawing instructions\". This means you can scale vector graphics up or downin reports or presentations without a loss of quality. Take a look for yourself:\n",
        "# In[22]:\n",
        "ax = plt.subplot()\n",
        "ax.scatter(y,ypred_calib_valid)\n",
        "lims = ax.get_xlim()\n",
        "ax.set_xlim(lims)\n",
        "ax.set_ylim(lims)\n",
        "ax.plot(lims,lims,ls=\"--\", c=\"grey\")\n",
        "ax.set_xlabel('Measured g')\n",
        "ax.set_ylabel('Modelled g')\n",
        "ax.set_title(f'Calibration-Validation Model Results: R$^2$ ={round(R2_calib_valid,2)}')\n",
        "plt.savefig('testf_pdf.pdf')\n",
        "plt.savefig('testf_png.png')\n",
        "# Open the two files and zoom in. You should notice that the .png is pixelated andthe .pdf file is not!\n",
        "# Generally, vector graphics are preferable. If they're not possible, make sure toincrease the dpi of the png file, e.g.\n",
        "#\n",
        "# plt.savefig(\"testf_png.png\",dpi = 400)\n",
        "#\n",
        "# All of these options are preferable to taking screenshots!\n",
        "# In[23]:\n",
        "print('correlation = ' + str(corr_coef[0, 1]))\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z1gM8gZSOIgn"
      },
      "source": [
        "## Best performer predictions"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_yhJ5b7C2MlT"
      },
      "source": [
        "### validation subset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "pJxB9pw0OLhe",
        "outputId": "ac96bae5-818d-48ad-a3c0-730bd7e8b3a7"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-9-4a07fb6c48a0>:5: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  all[var] = (all[var] - x_mean) / x_std\n",
            "<ipython-input-9-4a07fb6c48a0>:10: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  y_train[var] = (y_train[var] - y_mean) / y_std\n",
            "<ipython-input-9-4a07fb6c48a0>:12: SettingWithCopyWarning: \n",
            "A value is trying to be set on a copy of a slice from a DataFrame.\n",
            "Try using .loc[row_indexer,col_indexer] = value instead\n",
            "\n",
            "See the caveats in the documentation: https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
            "  all[var] = (all[var] - y_mean) / y_std\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "260.2766292232225\n",
            "260.2766292232225\n",
            "/content/drive/MyDrive/Colab Notebooks/EOSC510_Final/2b1fe551-20b7-45c2-911e-338208fcdd15.p\n",
            "   pred_tmin  pred_tmax\n",
            "0  -0.294195  -0.442747\n",
            "1  -0.157489  -0.333921\n",
            "2  -0.018235  -0.224372\n",
            "Index(['pred_tmin', 'pred_tmax'], dtype='object')\n",
            "2\n",
            "      Latitude  Longitude    tmin       Date    tmin_K        us    tmax  \\\n",
            "6608     48.71    -122.00 -5.6273 2023-04-15 -2.322553  0.903537  3.0756   \n",
            "6609     48.71    -121.96 -4.6361 2023-04-15 -2.163295  0.536719  5.3452   \n",
            "6610     48.71    -121.92 -3.9000 2023-04-15 -2.045024  0.194220  7.4603   \n",
            "\n",
            "        tmax_K  gfs_lon  gfs_lat  gfs_tmin  gfs_tmax  pred_tmin  pred_tmax  \n",
            "6608 -0.924240   -122.0    48.75 -1.527883 -1.516967  -0.294195  -0.442747  \n",
            "6609 -0.559578   -122.0    48.75 -1.527883 -1.516967  -0.157489  -0.333921  \n",
            "6610 -0.219740   -122.0    48.75 -1.527883 -1.516967  -0.018235  -0.224372  \n"
          ]
        }
      ],
      "source": [
        "# EVALUATE MODEL OUTPUT\n",
        "vars = [\"gfs_tmin\", \"gfs_tmax\", \"us\"]\n",
        "for var in vars:\n",
        "  test[var] = (test[var] - x_mean) / x_std\n",
        "  all[var] = (all[var] - x_mean) / x_std\n",
        "  val[var] = (val[var] - x_mean) / x_std\n",
        "\n",
        "y_vars = [\"tmin_K\", \"tmax_K\"]\n",
        "for var in y_vars:\n",
        "  y_train[var] = (y_train[var] - y_mean) / y_std\n",
        "  test[var] = (test[var] - y_mean) / y_std\n",
        "  all[var] = (all[var] - y_mean) / y_std\n",
        "  val[var] = (val[var] - y_mean) / y_std\n",
        "  print(y_mean)\n",
        "\n",
        "# VALIDATION\n",
        "# forecasted TMIN , Corrected TMIN, Actual TMIN\n",
        "predictions_df = get_predictions_from_best_performer(x_val)\n",
        "joined_data = join_data_to_original(predictions_df, val)\n",
        "\n",
        "joined_validation_predictions_filepath = f\"{wdir}/validation_data_and_predictions\"\n",
        "\n",
        "joined_data[\"point_error_tmin\"] = joined_data[\"pred_tmin\"] - joined_data[\"tmin_K\"]\n",
        "joined_data[\"point_error_tmax\"] = joined_data[\"pred_tmax\"] - joined_data[\"tmax_K\"]\n",
        "\n",
        "joined_data[\"gfs_error_tmin\"] = joined_data[\"pred_tmin\"] - joined_data[\"gfs_tmin\"]\n",
        "joined_data[\"gfs_error_tmax\"] = joined_data[\"pred_tmax\"] - joined_data[\"gfs_tmax\"]\n",
        "\n",
        "pickle.dump(joined_data,open(f\"{joined_validation_predictions_filepath}.p\",\"wb\"))\n",
        "joined_data.to_csv(f\"{joined_validation_predictions_filepath}.csv\")\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.figure(figsize=(12,8))\n",
        "\n",
        "plt.subplot(221)\n",
        "plt.scatter(range(len(nhn_best)),nhn_best)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('# Hidden Neurons')\n",
        "plt.title('Hidden Neurons')\n",
        "plt.subplot(222)\n",
        "plt.scatter(range(len(nhl_best)),nhl_best)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('# Hidden Layers')\n",
        "plt.title('Hidden Layers')\n",
        "plt.tight_layout()\n",
        "\n",
        "\n",
        "# plt.figure(figsize=(12,8))\n",
        "plt.scatter(len(RMSE_ensemble),RMSE_ensemble_mean,c='k',marker='*')\n",
        "plt.scatter(range(len(RMSE_ensemble)),RMSE_ensemble)\n",
        "plt.xlabel('Model #')\n",
        "plt.ylabel('RMSE')\n",
        "plt.title('Error')\n",
        "\n",
        "# plt.subplot(244)\n",
        "# plt.scatter(y_val,y_out_ensemble_mean)\n",
        "# plt.xlabel('y_test')\n",
        "# plt.ylabel('predicted y')\n",
        "# #plt.plot((np.min(y_test),np.max(y_test)),'k--')\n",
        "# # plt.xlabel('y_test')\n",
        "# # plt.ylabel('y_model')\n",
        "# # plt.title('Ensemble')\n",
        "# # plt.subplot(212)\n",
        "# # plt.plot(y_out_ensemble_mean, label='ensemble mean')\n",
        "# # plt.plot(np.array(y_val),alpha = 0.5, label='y_true')\n",
        "plt.tight_layout()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 655
        },
        "id": "apPFFB74YfH-",
        "outputId": "0bbf3c21-ab95-4bbe-fe59-02218caa2740"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "ValueError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-13-7ccf9c102e52>\u001b[0m in \u001b[0;36m<cell line: 17>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     15\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     16\u001b[0m \u001b[0;31m# plt.figure(figsize=(12,8))\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 17\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mRMSE_ensemble_mean\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'k'\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mmarker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'*'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     18\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mscatter\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrange\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mRMSE_ensemble\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mxlabel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'Model #'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/pyplot.py\u001b[0m in \u001b[0;36mscatter\u001b[0;34m(x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, edgecolors, plotnonfinite, data, **kwargs)\u001b[0m\n\u001b[1;32m   2860\u001b[0m         \u001b[0mvmin\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvmax\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlinewidths\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2861\u001b[0m         edgecolors=None, plotnonfinite=False, data=None, **kwargs):\n\u001b[0;32m-> 2862\u001b[0;31m     __ret = gca().scatter(\n\u001b[0m\u001b[1;32m   2863\u001b[0m         \u001b[0mx\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0ms\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0ms\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmarker\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mmarker\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcmap\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mcmap\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnorm\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mnorm\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2864\u001b[0m         \u001b[0mvmin\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvmin\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvmax\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mvmax\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0malpha\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0malpha\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mlinewidths\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mlinewidths\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/__init__.py\u001b[0m in \u001b[0;36minner\u001b[0;34m(ax, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1440\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0minner\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1441\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mdata\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1442\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mfunc\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0mmap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0msanitize_sequence\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1443\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1444\u001b[0m         \u001b[0mbound\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnew_sig\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mbind\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0max\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/axes/_axes.py\u001b[0m in \u001b[0;36mscatter\u001b[0;34m(self, x, y, s, c, marker, cmap, norm, vmin, vmax, alpha, linewidths, edgecolors, plotnonfinite, **kwargs)\u001b[0m\n\u001b[1;32m   4582\u001b[0m         \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mma\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mravel\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0my\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4583\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 4584\u001b[0;31m             \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"x and y must be the same size\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   4585\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   4586\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0ms\u001b[0m \u001b[0;32mis\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mValueError\u001b[0m: x and y must be the same size"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1200x800 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lH9g20xM2Xww",
        "outputId": "a887c157-3ccd-4ced-cc99-191b05b37b6a"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "/content/drive/MyDrive/Colab Notebooks/EOSC510_Final/2b1fe551-20b7-45c2-911e-338208fcdd15.p\n",
            "   pred_tmin  pred_tmax\n",
            "0   0.205577   0.108232\n",
            "1   0.115711   0.025037\n",
            "2   0.180193   0.087364\n",
            "Index(['pred_tmin', 'pred_tmax'], dtype='object')\n",
            "2\n",
            "     Latitude  Longitude    tmin       Date    tmin_K        us    tmax  \\\n",
            "209     48.96    -122.00 -2.5978 2023-04-15 -1.835797 -0.144225  6.9781   \n",
            "210     48.96    -121.96 -3.4048 2023-04-15 -1.965460  0.206380  5.6880   \n",
            "211     48.96    -121.92 -3.1109 2023-04-15 -1.918238 -0.048974  7.1849   \n",
            "\n",
            "       tmax_K  gfs_lon  gfs_lat  gfs_tmin  gfs_tmax  pred_tmin  pred_tmax  \n",
            "209 -0.297217   -122.0     49.0 -1.525777 -1.511901   0.205577   0.108232  \n",
            "210 -0.504500   -122.0     49.0 -1.525777 -1.511901   0.115711   0.025037  \n",
            "211 -0.263990   -122.0     49.0 -1.525777 -1.511901   0.180193   0.087364  \n",
            "/content/drive/MyDrive/Colab Notebooks/EOSC510_Final/2b1fe551-20b7-45c2-911e-338208fcdd15.p\n",
            "   pred_tmin  pred_tmax\n",
            "0   0.205577   0.108232\n",
            "1   0.115711   0.025037\n",
            "2   0.180193   0.087364\n",
            "Index(['pred_tmin', 'pred_tmax'], dtype='object')\n",
            "2\n",
            "     Latitude  Longitude    tmin       Date    tmin_K        us    tmax  \\\n",
            "209     48.96    -122.00 -2.5978 2023-04-15 -1.835797 -0.144225  6.9781   \n",
            "210     48.96    -121.96 -3.4048 2023-04-15 -1.965460  0.206380  5.6880   \n",
            "211     48.96    -121.92 -3.1109 2023-04-15 -1.918238 -0.048974  7.1849   \n",
            "\n",
            "       tmax_K  gfs_lon  gfs_lat  gfs_tmin  gfs_tmax  pred_tmin  pred_tmax  \n",
            "209 -0.297217   -122.0     49.0 -1.525777 -1.511901   0.205577   0.108232  \n",
            "210 -0.504500   -122.0     49.0 -1.525777 -1.511901   0.115711   0.025037  \n",
            "211 -0.263990   -122.0     49.0 -1.525777 -1.511901   0.180193   0.087364  \n"
          ]
        }
      ],
      "source": [
        "\n",
        "\n",
        "\n",
        "predictions_df = get_predictions_from_best_performer(x_tst)\n",
        "joined_data = join_data_to_original(predictions_df, test)\n",
        "\n",
        "joined_data[\"point_error_tmin\"] = joined_data[\"pred_tmin\"] - joined_data[\"tmin_K\"]\n",
        "joined_data[\"point_error_tmax\"] = joined_data[\"pred_tmax\"] - joined_data[\"tmax_K\"]\n",
        "\n",
        "joined_validation_predictions_filepath = f\"{wdir}/test_data_and_predictions_all\"\n",
        "\n",
        "pickle.dump(joined_data,open(f\"{joined_validation_predictions_filepath}.p\",\"wb\"))\n",
        "joined_data.to_csv(f\"{joined_validation_predictions_filepath}.csv\")\n",
        "\n",
        "\n",
        "# all\n",
        "predictions_df = get_predictions_from_best_performer(x_all)\n",
        "joined_data = join_data_to_original(predictions_df, all)\n",
        "\n",
        "joined_data[\"point_error_tmin\"] = joined_data[\"pred_tmin\"] - joined_data[\"tmin_K\"]\n",
        "joined_data[\"point_error_tmax\"] = joined_data[\"pred_tmax\"] - joined_data[\"tmax_K\"]\n",
        "\n",
        "joined_validation_predictions_filepath = f\"{wdir}/all_data_and_predictions_all\"\n",
        "\n",
        "pickle.dump(joined_data,open(f\"{joined_validation_predictions_filepath}.p\",\"wb\"))\n",
        "joined_data.to_csv(f\"{joined_validation_predictions_filepath}.csv\")\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "\n",
        "rmse(predictions_df, y_tst)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jTvXcjjDu7rU",
        "outputId": "26df326b-b740-4a6b-cfe0-ab29029f8c20"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "pred_tmax    0.0\n",
              "pred_tmin    0.0\n",
              "tmax_K       0.0\n",
              "tmin_K       0.0\n",
              "dtype: float64"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jsnzB0RWMhZ1",
        "outputId": "f320fa67-7773-4a32-8277-2498fb0b074c"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "261.2638188559237"
            ]
          },
          "execution_count": 27,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# tmin_K    0.285505\n",
        "# tmax_K    0.250797\n",
        "0.285505 * y_std + y_mean\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zCtx6iSCefGK"
      },
      "source": [
        "## plots (not in use)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 825
        },
        "id": "Li_A_WejPQBd",
        "outputId": "5d9128ff-f730-4728-d51a-63246193901f"
      },
      "outputs": [
        {
          "ename": "ValueError",
          "evalue": "ignored",
          "output_type": "error",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-16-924347043218>\u001b[0m in \u001b[0;36m<cell line: 5>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;31m# ensemble members\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_val\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtolist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredicted_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mT\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"k\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0malpha\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0.2\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      6\u001b[0m \u001b[0;31m# ensemble mean\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx_val\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtolist\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpredicted_output\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmean\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0maxis\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mc\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"b\"\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mlabel\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"ensemble mean\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/pyplot.py\u001b[0m in \u001b[0;36mplot\u001b[0;34m(scalex, scaley, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m   2810\u001b[0m \u001b[0;34m@\u001b[0m\u001b[0m_copy_docstring_and_deprecators\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mAxes\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mplot\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2811\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mplot\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscalex\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscaley\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2812\u001b[0;31m     return gca().plot(\n\u001b[0m\u001b[1;32m   2813\u001b[0m         \u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscalex\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mscalex\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mscaley\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mscaley\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2814\u001b[0m         **({\"data\": data} if data is not None else {}), **kwargs)\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/axes/_axes.py\u001b[0m in \u001b[0;36mplot\u001b[0;34m(self, scalex, scaley, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1686\u001b[0m         \"\"\"\n\u001b[1;32m   1687\u001b[0m         \u001b[0mkwargs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcbook\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mnormalize_kwargs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmlines\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mLine2D\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1688\u001b[0;31m         \u001b[0mlines\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_get_lines\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1689\u001b[0m         \u001b[0;32mfor\u001b[0m \u001b[0mline\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mlines\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1690\u001b[0m             \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madd_line\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mline\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, data, *args, **kwargs)\u001b[0m\n\u001b[1;32m    309\u001b[0m                 \u001b[0mthis\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    310\u001b[0m                 \u001b[0margs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 311\u001b[0;31m             yield from self._plot_args(\n\u001b[0m\u001b[1;32m    312\u001b[0m                 this, kwargs, ambiguous_fmt_datakey=ambiguous_fmt_datakey)\n\u001b[1;32m    313\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.10/dist-packages/matplotlib/axes/_base.py\u001b[0m in \u001b[0;36m_plot_args\u001b[0;34m(self, tup, kwargs, return_kwargs, ambiguous_fmt_datakey)\u001b[0m\n\u001b[1;32m    502\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    503\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m!=\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshape\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 504\u001b[0;31m             raise ValueError(f\"x and y must have same first dimension, but \"\n\u001b[0m\u001b[1;32m    505\u001b[0m                              f\"have shapes {x.shape} and {y.shape}\")\n\u001b[1;32m    506\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mx\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndim\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m2\u001b[0m \u001b[0;32mor\u001b[0m \u001b[0my\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mndim\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m2\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mValueError\u001b[0m: x and y must have same first dimension, but have shapes (432,) and (2, 432, 30)"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "\n",
        "# ensemble members\n",
        "plt.plot(x_val.index.tolist(), predicted_output.T,c=\"k\",alpha=0.2)\n",
        "# ensemble mean\n",
        "plt.plot(x_val.index.tolist(), predicted_output.mean(axis=0),c=\"b\",label=\"ensemble mean\")\n",
        "# the thing we are testing\n",
        "plt.plot(y_val,c=\"r\",label=\"y test\")\n",
        "plt.xlabel('t')\n",
        "plt.ylabel('predicted y')\n",
        "plt.legend()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "q-His5NNPIku"
      },
      "outputs": [],
      "source": [
        "# Validation\n",
        "\n",
        "\n",
        "# Test\n",
        "predicted_output = np.array([mod.predict(x_test) for mod in loaded_models])\n",
        "np.savetxt(f\"{wdir}/Westland.csv\", predicted_output.mean(axis=0), delimiter=\",\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TXZUlTbIq9Sq"
      },
      "source": [
        "Tmin error bigger than tmax -  maybe because GFS is coarse, it doesn't see the mountains. So that would mean it would consistently predict warmer temperatures. This explains why tmax error smaller than tmin error...  \n",
        "\n",
        "\n",
        "\n",
        "\n",
        "# Overview\n",
        "  Supervised MLP to bias correct HRRR forecasts over complex terrain.\n",
        "\n",
        "## Questions:\n",
        "- Can we improve the forecasts for a single station?\n",
        "    - input: forecast for station over time series, output: corrected forecast\n",
        "- Can we improve the TMin/TMax forecasts over stations with heavy orographic effects? (Washington state (cascades, rockies) vs BC (coastal range, rockies) )\n",
        "\n",
        "## Plan\n",
        "### Part 1 - single station\n",
        "1. Start with a single station time series\n",
        "2. Get a model that bias corrects for that station\n",
        "\n",
        "### Part 2 - forecast area bias correction\n",
        "1. Multiple inputs, multiple outputs (lat/lon grid)\n",
        "2. Use a CNN for this compared to MLP\n",
        "\n",
        "\n",
        "# Input\n",
        "  ## Forecasted Tmin Tmax values for a station. (single day)\n",
        "  ## Forecasted Tmin Tmax values for a station. (spectrum)\n",
        "\n",
        "# Output\n",
        "\n",
        "The output is a bias corrected forecast value for TMin, TMax.\n",
        "\n",
        "# Verification\n",
        "\n",
        "y_true will be the PRISM observations.\n",
        "\n",
        "\n",
        "# Methods\n",
        "\n",
        "MLP with outputs for TMin and TMax\n",
        "\n",
        "## Validation and training data\n",
        "\n",
        "Train on Washington stations\n",
        "Test on Vancouver if possible\n",
        "\n",
        "One month daily time series across Washington.\n",
        "Output is time series for one station."
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [
        "pGvCxhYgcnmC",
        "ZT2iKkmKUkd6",
        "_yhJ5b7C2MlT",
        "zCtx6iSCefGK"
      ],
      "provenance": [],
      "authorship_tag": "ABX9TyPG1TM3LMt0QeQOIYK5tdX6",
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}