Bad_Channels_Logistic_Regression.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Logistic Regression Model for Predicting Bad Hydrophone Channels**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The scope of this project is to properly model and train an algorithm that is able to identify bad receivers.\n",
    "\n",
    "The data contains different receiver statistics which include some instrumentational feedback (capacitance, leakage, sensitivity, etc..) decoded from original SEGD field data. There is also some computed statistics such as, root mean square of (noise) amplitude, standard deviation of RMS, etc.., which would traditionally be used to \"pick bad channels\") and also some user defined target variable (identified bad receivers)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import csv\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import seaborn as sns; sns.set()\n",
    "\n",
    "from pandas import DataFrame \n",
    "from sklearn.preprocessing import StandardScaler, normalize\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.mixture import GaussianMixture \n",
    "from sklearn.metrics import silhouette_score\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import metrics\n",
    "import matplotlib.ticker as ticker\n",
    "from sklearn import preprocessing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "List files in directory to use:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "['.git', '001_amp_maps_2890.csv', '002_amp_maps_3422.csv', '003_amp_maps_2918.csv', '004_amp_maps_3450.csv', '005_amp_maps_2946.csv', '006_amp_maps_3478.csv', '008_P_amp_maps_2265P01.csv', 'Bad_Channels.ipynb', 'Bad_Channels_Logistic_Regression.ipynb', 'decode_listing_features_001.csv', 'decode_listing_features_002.csv', 'decode_listing_features_003.csv', 'decode_listing_features_003_old.csv', 'decode_listing_features_004.csv', 'decode_listing_features_005.csv', 'decode_listing_features_006.csv', 'decode_listing_features_010_old.csv', 'Log_ROC.png', 'README.md', 'RMS.png']\n"
     ]
    }
   ],
   "source": [
    "import sys, os\n",
    "print(os.listdir())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we import the first set of data, which include have some sequence specific features, including the target feature \"BAD_CHAN_FLAG\" (binary, 0 good, 1 bad)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "stats1=pd.read_csv('decode_listing_features_001.csv')\n",
    "stats1['SEQ']=1\n",
    "#stats1a=pd.read_csv('decode_listing_features_002.csv')\n",
    "#stats1a['SEQ']=2\n",
    "#stats1b=pd.read_csv('decode_listing_features_003.csv')\n",
    "#stats1b['SEQ']=3\n",
    "#stats1c=pd.read_csv('decode_listing_features_004.csv')\n",
    "#stats1c['SEQ']=4\n",
    "#stats1d=pd.read_csv('decode_listing_features_005.csv')\n",
    "#stats1d['SEQ']=5\n",
    "#stats1e=pd.read_csv('decode_listing_features_005.csv')\n",
    "#stats1e['SEQ']=6\n",
    "#stats1=stats1.append(stats1a)\n",
    "#stats1=stats1.append(stats1b)\n",
    "#stats1=stats1.append(stats1c)\n",
    "#stats1=stats1.append(stats1d)\n",
    "#stats1=stats1.append(stats1e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>BAD_CHAN_FLAG</th>\n",
       "      <th>CAP_CUT_ERR</th>\n",
       "      <th>CAP_ERR</th>\n",
       "      <th>CAP_VAL</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS</th>\n",
       "      <th>DEVICE</th>\n",
       "      <th>LEAK_ERR</th>\n",
       "      <th>LEAK_VAL</th>\n",
       "      <th>MARG_CHAN_FLAG</th>\n",
       "      <th>SENS</th>\n",
       "      <th>SEQ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.691345</td>\n",
       "      <td>1</td>\n",
       "      <td>10.58</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.66</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.366440</td>\n",
       "      <td>2</td>\n",
       "      <td>7.28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.22</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>234.802597</td>\n",
       "      <td>3</td>\n",
       "      <td>6.10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.32</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233.590988</td>\n",
       "      <td>4</td>\n",
       "      <td>6.78</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.26</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.584885</td>\n",
       "      <td>5</td>\n",
       "      <td>7.02</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   BAD_CHAN_FLAG  CAP_CUT_ERR  CAP_ERR     CAP_VAL  CHANNEL  DEEP_RMS  DEVICE  \\\n",
       "0              0            0        0  232.691345        1     10.58       4   \n",
       "1              0            0        0  228.366440        2      7.28       0   \n",
       "2              0            0        0  234.802597        3      6.10       0   \n",
       "3              0            0        0  233.590988        4      6.78       0   \n",
       "4              0            0        0  228.584885        5      7.02       0   \n",
       "\n",
       "   LEAK_ERR  LEAK_VAL  MARG_CHAN_FLAG   SENS  SEQ  \n",
       "0         0       5.0               0  31.66    1  \n",
       "1         0       5.0               0  31.22    1  \n",
       "2         0       5.0               0  30.32    1  \n",
       "3         0       5.0               0  31.26    1  \n",
       "4         0       5.0               0  31.13    1  "
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats1.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we import the RMS (noise) amplitude maps that will be used to derive more statistics. These provide an overview of recorded noise on the streamers - both backround and in our case when individual components breakdown and show up as bad channels."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "stats2=pd.read_csv('001_amp_maps_2890.csv')\n",
    "stats2['SEQ']=1\n",
    "#stats2=pd.read_csv('008_P_amp_maps_2265P01.csv')\n",
    "#stats2['SEQ']=1\n",
    "#stats2a=pd.read_csv('002_amp_maps_3422.csv')\n",
    "#stats2a['SEQ']=2\n",
    "#stats2b=pd.read_csv('003_amp_maps_2918.csv')\n",
    "#stats2b['SEQ']=3\n",
    "#stats2c=pd.read_csv('004_amp_maps_3450.csv')\n",
    "#stats2c['SEQ']=4\n",
    "#stats2d=pd.read_csv('005_amp_maps_2946.csv')\n",
    "#stats2d['SEQ']=5\n",
    "#stats2e=pd.read_csv('006_amp_maps_3478.csv')#\n",
    "#stats2e['SEQ']=6"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Due to large file size only run more than one sequence and the following append code block if the computer has enough RAM."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#stats2=stats2.append(stats2a)\n",
    "#stats2=stats2.append(stats2b)\n",
    "#stats2=stats2.append(stats2c)\n",
    "#stats2=stats2.append(stats2d)\n",
    "#stats2=stats2.append(stats2e)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SRC_POINT</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>WC_RMS</th>\n",
       "      <th>TGT_RMS</th>\n",
       "      <th>DEEP_RMS</th>\n",
       "      <th>FFID</th>\n",
       "      <th>CABLE</th>\n",
       "      <th>CABTR</th>\n",
       "      <th>SEQ</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5071</td>\n",
       "      <td>1</td>\n",
       "      <td>5.415281</td>\n",
       "      <td>30.081125</td>\n",
       "      <td>12.606074</td>\n",
       "      <td>1001</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5071</td>\n",
       "      <td>2</td>\n",
       "      <td>6.816298</td>\n",
       "      <td>28.298389</td>\n",
       "      <td>9.360264</td>\n",
       "      <td>1001</td>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>5071</td>\n",
       "      <td>3</td>\n",
       "      <td>6.319650</td>\n",
       "      <td>25.300581</td>\n",
       "      <td>8.873336</td>\n",
       "      <td>1001</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5071</td>\n",
       "      <td>4</td>\n",
       "      <td>2.712788</td>\n",
       "      <td>28.077599</td>\n",
       "      <td>7.638037</td>\n",
       "      <td>1001</td>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5071</td>\n",
       "      <td>5</td>\n",
       "      <td>4.047064</td>\n",
       "      <td>27.413845</td>\n",
       "      <td>9.036798</td>\n",
       "      <td>1001</td>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SRC_POINT  CHANNEL    WC_RMS    TGT_RMS   DEEP_RMS  FFID  CABLE  CABTR  SEQ\n",
       "0       5071        1  5.415281  30.081125  12.606074  1001      1      1    1\n",
       "1       5071        2  6.816298  28.298389   9.360264  1001      1      2    1\n",
       "2       5071        3  6.319650  25.300581   8.873336  1001      1      3    1\n",
       "3       5071        4  2.712788  28.077599   7.638037  1001      1      4    1\n",
       "4       5071        5  4.047064  27.413845   9.036798  1001      1      5    1"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stats2.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Get standard deviation of root mean squared values, grouped by sequence and receiver number."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_=stats2.groupby(['SEQ','CHANNEL'])['DEEP_RMS'].agg(np.std)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_=pd.DataFrame(std_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "std_.sort_values(['SEQ','CHANNEL'],inplace=True)\n",
    "stats1.sort_values(['SEQ','CHANNEL'],inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9072"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std_.reset_index(inplace=True)\n",
    "len(std_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9072"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(stats1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SEQ</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3.725214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.094594</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.885017</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2.093174</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2.046160</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SEQ  CHANNEL  DEEP_RMS\n",
       "0    1        1  3.725214\n",
       "1    1        2  2.094594\n",
       "2    1        3  1.885017\n",
       "3    1        4  2.093174\n",
       "4    1        5  2.046160"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "std_.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Merge the two dataframes."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "df1=pd.merge(std_,stats1 , how='left', left_on=['SEQ','CHANNEL'], right_on=['SEQ','CHANNEL'])\n",
    "df2=df1.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SEQ</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS_x</th>\n",
       "      <th>BAD_CHAN_FLAG</th>\n",
       "      <th>CAP_CUT_ERR</th>\n",
       "      <th>CAP_ERR</th>\n",
       "      <th>CAP_VAL</th>\n",
       "      <th>DEEP_RMS_y</th>\n",
       "      <th>DEVICE</th>\n",
       "      <th>LEAK_ERR</th>\n",
       "      <th>LEAK_VAL</th>\n",
       "      <th>MARG_CHAN_FLAG</th>\n",
       "      <th>SENS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3.725214</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.691345</td>\n",
       "      <td>10.58</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.094594</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.366440</td>\n",
       "      <td>7.28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.885017</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>234.802597</td>\n",
       "      <td>6.10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2.093174</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233.590988</td>\n",
       "      <td>6.78</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2.046160</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.584885</td>\n",
       "      <td>7.02</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1.960650</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.264236</td>\n",
       "      <td>6.56</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>2.006694</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.004303</td>\n",
       "      <td>6.84</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>2.026388</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>231.898392</td>\n",
       "      <td>6.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>1.988935</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.703339</td>\n",
       "      <td>7.01</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>2.097428</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>226.584427</td>\n",
       "      <td>7.53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SEQ  CHANNEL  DEEP_RMS_x  BAD_CHAN_FLAG  CAP_CUT_ERR  CAP_ERR     CAP_VAL  \\\n",
       "0    1        1    3.725214              0            0        0  232.691345   \n",
       "1    1        2    2.094594              0            0        0  228.366440   \n",
       "2    1        3    1.885017              0            0        0  234.802597   \n",
       "3    1        4    2.093174              0            0        0  233.590988   \n",
       "4    1        5    2.046160              0            0        0  228.584885   \n",
       "5    1        6    1.960650              0            0        0  230.264236   \n",
       "6    1        7    2.006694              0            0        0  230.004303   \n",
       "7    1        8    2.026388              0            0        0  231.898392   \n",
       "8    1        9    1.988935              0            0        0  230.703339   \n",
       "9    1       10    2.097428              0            0        0  226.584427   \n",
       "\n",
       "   DEEP_RMS_y  DEVICE  LEAK_ERR  LEAK_VAL  MARG_CHAN_FLAG   SENS  \n",
       "0       10.58       4         0       5.0               0  31.66  \n",
       "1        7.28       0         0       5.0               0  31.22  \n",
       "2        6.10       0         0       5.0               0  30.32  \n",
       "3        6.78       0         0       5.0               0  31.26  \n",
       "4        7.02       0         0       5.0               0  31.13  \n",
       "5        6.56       0         0       5.0               0  30.82  \n",
       "6        6.84       0         0       5.0               0  31.01  \n",
       "7        6.79       0         0       5.0               0  31.13  \n",
       "8        7.01       0         0       5.0               0  30.98  \n",
       "9        7.53       0         0       5.0               0  31.27  "
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df2.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Plot the Root Mean Square (RMS) of (noise) Amplitudes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "a=stats2['TGT_RMS']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "b=a.to_list()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Re-shape the list for plotting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "c=np.reshape(b,(-1,9072))\n",
    "c=np.transpose(c)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A little clipping of the value to 100 for color scaling."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "c=np.clip(c,0,100)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.colors as colly\n",
    "\n",
    "cmap1= colly.LinearSegmentedColormap.from_list(\"\" , ['blue','green','red'])\n",
    "\n",
    "im=plt.imshow(c, cmap=cmap1)\n",
    "plt.colorbar(im)\n",
    "plt.rcParams[\"figure.figsize\"] = (70,20)\n",
    "plt.savefig('RMS.png')\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Now for the modelling part...**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SEQ</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS_x</th>\n",
       "      <th>BAD_CHAN_FLAG</th>\n",
       "      <th>CAP_CUT_ERR</th>\n",
       "      <th>CAP_ERR</th>\n",
       "      <th>CAP_VAL</th>\n",
       "      <th>DEEP_RMS_y</th>\n",
       "      <th>DEVICE</th>\n",
       "      <th>LEAK_ERR</th>\n",
       "      <th>LEAK_VAL</th>\n",
       "      <th>MARG_CHAN_FLAG</th>\n",
       "      <th>SENS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3.725214</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.691345</td>\n",
       "      <td>10.58</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.094594</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.366440</td>\n",
       "      <td>7.28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.885017</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>234.802597</td>\n",
       "      <td>6.10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2.093174</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233.590988</td>\n",
       "      <td>6.78</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2.046160</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.584885</td>\n",
       "      <td>7.02</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>6</td>\n",
       "      <td>1.960650</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.264236</td>\n",
       "      <td>6.56</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.82</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1</td>\n",
       "      <td>7</td>\n",
       "      <td>2.006694</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.004303</td>\n",
       "      <td>6.84</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1</td>\n",
       "      <td>8</td>\n",
       "      <td>2.026388</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>231.898392</td>\n",
       "      <td>6.79</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>1</td>\n",
       "      <td>9</td>\n",
       "      <td>1.988935</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>230.703339</td>\n",
       "      <td>7.01</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.98</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1</td>\n",
       "      <td>10</td>\n",
       "      <td>2.097428</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>226.584427</td>\n",
       "      <td>7.53</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.27</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SEQ  CHANNEL  DEEP_RMS_x  BAD_CHAN_FLAG  CAP_CUT_ERR  CAP_ERR     CAP_VAL  \\\n",
       "0    1        1    3.725214              0            0        0  232.691345   \n",
       "1    1        2    2.094594              0            0        0  228.366440   \n",
       "2    1        3    1.885017              0            0        0  234.802597   \n",
       "3    1        4    2.093174              0            0        0  233.590988   \n",
       "4    1        5    2.046160              0            0        0  228.584885   \n",
       "5    1        6    1.960650              0            0        0  230.264236   \n",
       "6    1        7    2.006694              0            0        0  230.004303   \n",
       "7    1        8    2.026388              0            0        0  231.898392   \n",
       "8    1        9    1.988935              0            0        0  230.703339   \n",
       "9    1       10    2.097428              0            0        0  226.584427   \n",
       "\n",
       "   DEEP_RMS_y  DEVICE  LEAK_ERR  LEAK_VAL  MARG_CHAN_FLAG   SENS  \n",
       "0       10.58       4         0       5.0               0  31.66  \n",
       "1        7.28       0         0       5.0               0  31.22  \n",
       "2        6.10       0         0       5.0               0  30.32  \n",
       "3        6.78       0         0       5.0               0  31.26  \n",
       "4        7.02       0         0       5.0               0  31.13  \n",
       "5        6.56       0         0       5.0               0  30.82  \n",
       "6        6.84       0         0       5.0               0  31.01  \n",
       "7        6.79       0         0       5.0               0  31.13  \n",
       "8        7.01       0         0       5.0               0  30.98  \n",
       "9        7.53       0         0       5.0               0  31.27  "
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.head(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Drop one by one the features we do not need..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_16856\\1774694302.py:1: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  df1=df1.drop('CAP_CUT_ERR',1)\n",
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_16856\\1774694302.py:2: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  df1=df1.drop('CAP_ERR',1)\n",
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_16856\\1774694302.py:3: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  df1=df1.drop('LEAK_ERR',1)\n",
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_16856\\1774694302.py:4: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  df1=df1.drop('SENS',1)\n",
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_16856\\1774694302.py:5: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  df1=df1.drop('MARG_CHAN_FLAG',1)\n"
     ]
    }
   ],
   "source": [
    "df1=df1.drop('CAP_CUT_ERR',1)\n",
    "df1=df1.drop('CAP_ERR',1)\n",
    "df1=df1.drop('LEAK_ERR',1)\n",
    "df1=df1.drop('SENS',1)\n",
    "df1=df1.drop('MARG_CHAN_FLAG',1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>SEQ</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS_x</th>\n",
       "      <th>BAD_CHAN_FLAG</th>\n",
       "      <th>CAP_CUT_ERR</th>\n",
       "      <th>CAP_ERR</th>\n",
       "      <th>CAP_VAL</th>\n",
       "      <th>DEEP_RMS_y</th>\n",
       "      <th>DEVICE</th>\n",
       "      <th>LEAK_ERR</th>\n",
       "      <th>LEAK_VAL</th>\n",
       "      <th>MARG_CHAN_FLAG</th>\n",
       "      <th>SENS</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>3.725214</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.691345</td>\n",
       "      <td>10.58</td>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.66</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>2.094594</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.366440</td>\n",
       "      <td>7.28</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "      <td>1.885017</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>234.802597</td>\n",
       "      <td>6.10</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>4</td>\n",
       "      <td>2.093174</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>233.590988</td>\n",
       "      <td>6.78</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>5</td>\n",
       "      <td>2.046160</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>228.584885</td>\n",
       "      <td>7.02</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.13</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   SEQ  CHANNEL  DEEP_RMS_x  BAD_CHAN_FLAG  CAP_CUT_ERR  CAP_ERR     CAP_VAL  \\\n",
       "0    1        1    3.725214              0            0        0  232.691345   \n",
       "1    1        2    2.094594              0            0        0  228.366440   \n",
       "2    1        3    1.885017              0            0        0  234.802597   \n",
       "3    1        4    2.093174              0            0        0  233.590988   \n",
       "4    1        5    2.046160              0            0        0  228.584885   \n",
       "\n",
       "   DEEP_RMS_y  DEVICE  LEAK_ERR  LEAK_VAL  MARG_CHAN_FLAG   SENS  \n",
       "0       10.58       4         0       5.0               0  31.66  \n",
       "1        7.28       0         0       5.0               0  31.22  \n",
       "2        6.10       0         0       5.0               0  30.32  \n",
       "3        6.78       0         0       5.0               0  31.26  \n",
       "4        7.02       0         0       5.0               0  31.13  "
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_12412\\1240260745.py:3: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  X=df1.drop(['BAD_CHAN_FLAG'],1)\n"
     ]
    }
   ],
   "source": [
    "pd.to_numeric(df1['BAD_CHAN_FLAG'])\n",
    "y=pd.DataFrame(df1['BAD_CHAN_FLAG'])\n",
    "X=df1.drop(['BAD_CHAN_FLAG'],1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "yy=y.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\sirproc\\AppData\\Local\\Temp\\ipykernel_12412\\447026479.py:2: FutureWarning: In a future version of pandas all arguments of DataFrame.drop except for the argument 'labels' will be keyword-only.\n",
      "  X2=X2.drop(['CHANNEL','SEQ'],1)\n"
     ]
    }
   ],
   "source": [
    "X2=X.copy()\n",
    "X2=X2.drop(['CHANNEL','SEQ'],1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate the correlation values\n",
    "feature_cols = X2.columns\n",
    "corr_values = X2[feature_cols].corr()\n",
    "\n",
    "# Simplify by emptying all the data below the diagonal\n",
    "tril_index = np.tril_indices_from(corr_values)\n",
    "\n",
    "# Make the unused values NaNs\n",
    "for coord in zip(*tril_index):\n",
    "    corr_values.iloc[coord[0], coord[1]] = np.NaN\n",
    "    \n",
    "# Stack the data and convert to a data frame\n",
    "corr_values = (corr_values\n",
    "               .stack()\n",
    "               .to_frame()\n",
    "               .reset_index()\n",
    "               .rename(columns={'level_0':'feature1',\n",
    "                                'level_1':'feature2',\n",
    "                                0:'correlation'}))\n",
    "\n",
    "# Get the absolute values for sorting\n",
    "corr_values['abs_correlation'] = corr_values.correlation.abs()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x576 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "sns.set_context('talk')\n",
    "sns.set_style('white')\n",
    "\n",
    "ax = corr_values.abs_correlation.hist(bins=10, figsize=(12, 8))\n",
    "ax.set(xlabel='Absolute Correlation', ylabel='Frequency');"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>feature1</th>\n",
       "      <th>feature2</th>\n",
       "      <th>correlation</th>\n",
       "      <th>abs_correlation</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>DEEP_RMS_y</td>\n",
       "      <td>SENS</td>\n",
       "      <td>0.723199</td>\n",
       "      <td>0.723199</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>DEEP_RMS_x</td>\n",
       "      <td>SENS</td>\n",
       "      <td>0.562433</td>\n",
       "      <td>0.562433</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>DEEP_RMS_x</td>\n",
       "      <td>DEEP_RMS_y</td>\n",
       "      <td>0.554213</td>\n",
       "      <td>0.554213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>LEAK_ERR</td>\n",
       "      <td>MARG_CHAN_FLAG</td>\n",
       "      <td>0.485388</td>\n",
       "      <td>0.485388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>DEEP_RMS_y</td>\n",
       "      <td>DEVICE</td>\n",
       "      <td>0.253006</td>\n",
       "      <td>0.253006</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>LEAK_VAL</td>\n",
       "      <td>MARG_CHAN_FLAG</td>\n",
       "      <td>-0.450038</td>\n",
       "      <td>0.450038</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>LEAK_ERR</td>\n",
       "      <td>LEAK_VAL</td>\n",
       "      <td>-0.889073</td>\n",
       "      <td>0.889073</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      feature1        feature2  correlation  abs_correlation\n",
       "17  DEEP_RMS_y            SENS     0.723199         0.723199\n",
       "6   DEEP_RMS_x            SENS     0.562433         0.562433\n",
       "1   DEEP_RMS_x      DEEP_RMS_y     0.554213         0.554213\n",
       "23    LEAK_ERR  MARG_CHAN_FLAG     0.485388         0.485388\n",
       "13  DEEP_RMS_y          DEVICE     0.253006         0.253006\n",
       "25    LEAK_VAL  MARG_CHAN_FLAG    -0.450038         0.450038\n",
       "22    LEAK_ERR        LEAK_VAL    -0.889073         0.889073"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# The most highly correlated values\n",
    "corr_values.sort_values('correlation', ascending=False).query('abs_correlation>0.2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Logistic Regression**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX=X2.values.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[3.725214248963877, 0.0, 0.0, 232.691345, 10.58, 4.0, 0.0, 5.0, 0.0, 31.66],\n",
       " [2.0945937529968033, 0.0, 0.0, 228.36644, 7.28, 0.0, 0.0, 5.0, 0.0, 31.22],\n",
       " [1.8850165095194575, 0.0, 0.0, 234.802597, 6.1, 0.0, 0.0, 5.0, 0.0, 30.32],\n",
       " [2.0931738147193406, 0.0, 0.0, 233.590988, 6.78, 0.0, 0.0, 5.0, 0.0, 31.26],\n",
       " [2.0461598423613534, 0.0, 0.0, 228.584885, 7.02, 0.0, 0.0, 5.0, 0.0, 31.13]]"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "XX[0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We use a standard scalar to normalise the dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "XX=preprocessing.StandardScaler().fit(XX).transform(XX)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 3.28383591e+00,  0.00000000e+00,  0.00000000e+00,\n",
       "        -9.74047382e-02,  2.43199616e+00,  2.87776781e+00,\n",
       "        -5.15026203e-02,  6.34679145e-02, -1.06106077e-01,\n",
       "         9.91887362e-01],\n",
       "       [ 5.46978559e-01,  0.00000000e+00,  0.00000000e+00,\n",
       "        -5.63734644e-01,  3.87188179e-01, -2.90955285e-01,\n",
       "        -5.15026203e-02,  6.34679145e-02, -1.06106077e-01,\n",
       "         8.59420139e-01],\n",
       "       [ 1.95221041e-01,  0.00000000e+00,  0.00000000e+00,\n",
       "         1.30239556e-01, -3.43985586e-01, -2.90955285e-01,\n",
       "        -5.15026203e-02,  6.34679145e-02, -1.06106077e-01,\n",
       "         5.88464454e-01],\n",
       "       [ 5.44595314e-01,  0.00000000e+00,  0.00000000e+00,\n",
       "        -4.01350614e-04,  7.73687870e-02, -2.90955285e-01,\n",
       "        -5.15026203e-02,  6.34679145e-02, -1.06106077e-01,\n",
       "         8.71462614e-01],\n",
       "       [ 4.65686374e-01,  0.00000000e+00,  0.00000000e+00,\n",
       "        -5.40180963e-01,  2.26082095e-01, -2.90955285e-01,\n",
       "        -5.15026203e-02,  6.34679145e-02, -1.06106077e-01,\n",
       "         8.32324570e-01]])"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "XX[0:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Test the logistic regression on the whole dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Optimization terminated successfully.\n",
      "         Current function value: 0.692296\n",
      "         Iterations 4\n"
     ]
    },
    {
     "ename": "LinAlgError",
     "evalue": "Singular matrix",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mLinAlgError\u001b[0m                               Traceback (most recent call last)",
      "\u001b[1;32mz:\\Dan\\Code\\Seismic\\Bad_Channels_Logistic_Regression.ipynb Cell 51'\u001b[0m in \u001b[0;36m<cell line: 3>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000034?line=0'>1</a>\u001b[0m \u001b[39mimport\u001b[39;00m \u001b[39mstatsmodels\u001b[39;00m\u001b[39m.\u001b[39;00m\u001b[39mapi\u001b[39;00m \u001b[39mas\u001b[39;00m \u001b[39msm\u001b[39;00m\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000034?line=1'>2</a>\u001b[0m logit_model\u001b[39m=\u001b[39msm\u001b[39m.\u001b[39mLogit(yy,XX)\n\u001b[1;32m----> <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000034?line=2'>3</a>\u001b[0m result\u001b[39m=\u001b[39mlogit_model\u001b[39m.\u001b[39;49mfit()\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000034?line=3'>4</a>\u001b[0m \u001b[39mprint\u001b[39m(result\u001b[39m.\u001b[39msummary2())\n",
      "File \u001b[1;32mc:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\statsmodels\\discrete\\discrete_model.py:1983\u001b[0m, in \u001b[0;36mLogit.fit\u001b[1;34m(self, start_params, method, maxiter, full_output, disp, callback, **kwargs)\u001b[0m\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1979'>1980</a>\u001b[0m \u001b[39m@Appender\u001b[39m(DiscreteModel\u001b[39m.\u001b[39mfit\u001b[39m.\u001b[39m\u001b[39m__doc__\u001b[39m)\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1980'>1981</a>\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39mfit\u001b[39m(\u001b[39mself\u001b[39m, start_params\u001b[39m=\u001b[39m\u001b[39mNone\u001b[39;00m, method\u001b[39m=\u001b[39m\u001b[39m'\u001b[39m\u001b[39mnewton\u001b[39m\u001b[39m'\u001b[39m, maxiter\u001b[39m=\u001b[39m\u001b[39m35\u001b[39m,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1981'>1982</a>\u001b[0m         full_output\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m, disp\u001b[39m=\u001b[39m\u001b[39m1\u001b[39m, callback\u001b[39m=\u001b[39m\u001b[39mNone\u001b[39;00m, \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs):\n\u001b[1;32m-> <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1982'>1983</a>\u001b[0m     bnryfit \u001b[39m=\u001b[39m \u001b[39msuper\u001b[39m()\u001b[39m.\u001b[39mfit(start_params\u001b[39m=\u001b[39mstart_params,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1983'>1984</a>\u001b[0m                           method\u001b[39m=\u001b[39mmethod,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1984'>1985</a>\u001b[0m                           maxiter\u001b[39m=\u001b[39mmaxiter,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1985'>1986</a>\u001b[0m                           full_output\u001b[39m=\u001b[39mfull_output,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1986'>1987</a>\u001b[0m                           disp\u001b[39m=\u001b[39mdisp,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1987'>1988</a>\u001b[0m                           callback\u001b[39m=\u001b[39mcallback,\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1988'>1989</a>\u001b[0m                           \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs)\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1990'>1991</a>\u001b[0m     discretefit \u001b[39m=\u001b[39m LogitResults(\u001b[39mself\u001b[39m, bnryfit)\n\u001b[0;32m   <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=1991'>1992</a>\u001b[0m     \u001b[39mreturn\u001b[39;00m BinaryResultsWrapper(discretefit)\n",
      "File \u001b[1;32mc:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\statsmodels\\discrete\\discrete_model.py:230\u001b[0m, in \u001b[0;36mDiscreteModel.fit\u001b[1;34m(self, start_params, method, maxiter, full_output, disp, callback, **kwargs)\u001b[0m\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=226'>227</a>\u001b[0m \u001b[39melse\u001b[39;00m:\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=227'>228</a>\u001b[0m     \u001b[39mpass\u001b[39;00m  \u001b[39m# TODO: make a function factory to have multiple call-backs\u001b[39;00m\n\u001b[1;32m--> <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=229'>230</a>\u001b[0m mlefit \u001b[39m=\u001b[39m \u001b[39msuper\u001b[39m()\u001b[39m.\u001b[39mfit(start_params\u001b[39m=\u001b[39mstart_params,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=230'>231</a>\u001b[0m                      method\u001b[39m=\u001b[39mmethod,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=231'>232</a>\u001b[0m                      maxiter\u001b[39m=\u001b[39mmaxiter,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=232'>233</a>\u001b[0m                      full_output\u001b[39m=\u001b[39mfull_output,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=233'>234</a>\u001b[0m                      disp\u001b[39m=\u001b[39mdisp,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=234'>235</a>\u001b[0m                      callback\u001b[39m=\u001b[39mcallback,\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=235'>236</a>\u001b[0m                      \u001b[39m*\u001b[39m\u001b[39m*\u001b[39mkwargs)\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/discrete/discrete_model.py?line=237'>238</a>\u001b[0m \u001b[39mreturn\u001b[39;00m mlefit\n",
      "File \u001b[1;32mc:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\statsmodels\\base\\model.py:579\u001b[0m, in \u001b[0;36mLikelihoodModel.fit\u001b[1;34m(self, start_params, method, maxiter, full_output, disp, fargs, callback, retall, skip_hessian, **kwargs)\u001b[0m\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/base/model.py?line=576'>577</a>\u001b[0m     Hinv \u001b[39m=\u001b[39m cov_params_func(\u001b[39mself\u001b[39m, xopt, retvals)\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/base/model.py?line=577'>578</a>\u001b[0m \u001b[39melif\u001b[39;00m method \u001b[39m==\u001b[39m \u001b[39m'\u001b[39m\u001b[39mnewton\u001b[39m\u001b[39m'\u001b[39m \u001b[39mand\u001b[39;00m full_output:\n\u001b[1;32m--> <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/base/model.py?line=578'>579</a>\u001b[0m     Hinv \u001b[39m=\u001b[39m np\u001b[39m.\u001b[39;49mlinalg\u001b[39m.\u001b[39;49minv(\u001b[39m-\u001b[39;49mretvals[\u001b[39m'\u001b[39;49m\u001b[39mHessian\u001b[39;49m\u001b[39m'\u001b[39;49m]) \u001b[39m/\u001b[39m nobs\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/base/model.py?line=579'>580</a>\u001b[0m \u001b[39melif\u001b[39;00m \u001b[39mnot\u001b[39;00m skip_hessian:\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/statsmodels/base/model.py?line=580'>581</a>\u001b[0m     H \u001b[39m=\u001b[39m \u001b[39m-\u001b[39m\u001b[39m1\u001b[39m \u001b[39m*\u001b[39m \u001b[39mself\u001b[39m\u001b[39m.\u001b[39mhessian(xopt)\n",
      "File \u001b[1;32m<__array_function__ internals>:180\u001b[0m, in \u001b[0;36minv\u001b[1;34m(*args, **kwargs)\u001b[0m\n",
      "File \u001b[1;32mc:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\numpy\\linalg\\linalg.py:545\u001b[0m, in \u001b[0;36minv\u001b[1;34m(a)\u001b[0m\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=542'>543</a>\u001b[0m signature \u001b[39m=\u001b[39m \u001b[39m'\u001b[39m\u001b[39mD->D\u001b[39m\u001b[39m'\u001b[39m \u001b[39mif\u001b[39;00m isComplexType(t) \u001b[39melse\u001b[39;00m \u001b[39m'\u001b[39m\u001b[39md->d\u001b[39m\u001b[39m'\u001b[39m\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=543'>544</a>\u001b[0m extobj \u001b[39m=\u001b[39m get_linalg_error_extobj(_raise_linalgerror_singular)\n\u001b[1;32m--> <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=544'>545</a>\u001b[0m ainv \u001b[39m=\u001b[39m _umath_linalg\u001b[39m.\u001b[39;49minv(a, signature\u001b[39m=\u001b[39;49msignature, extobj\u001b[39m=\u001b[39;49mextobj)\n\u001b[0;32m    <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=545'>546</a>\u001b[0m \u001b[39mreturn\u001b[39;00m wrap(ainv\u001b[39m.\u001b[39mastype(result_t, copy\u001b[39m=\u001b[39m\u001b[39mFalse\u001b[39;00m))\n",
      "File \u001b[1;32mc:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\numpy\\linalg\\linalg.py:88\u001b[0m, in \u001b[0;36m_raise_linalgerror_singular\u001b[1;34m(err, flag)\u001b[0m\n\u001b[0;32m     <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=86'>87</a>\u001b[0m \u001b[39mdef\u001b[39;00m \u001b[39m_raise_linalgerror_singular\u001b[39m(err, flag):\n\u001b[1;32m---> <a href='file:///c%3A/Users/sirproc/Anaconda3/envs/data_science/lib/site-packages/numpy/linalg/linalg.py?line=87'>88</a>\u001b[0m     \u001b[39mraise\u001b[39;00m LinAlgError(\u001b[39m\"\u001b[39m\u001b[39mSingular matrix\u001b[39m\u001b[39m\"\u001b[39m)\n",
      "\u001b[1;31mLinAlgError\u001b[0m: Singular matrix"
     ]
    }
   ],
   "source": [
    "import statsmodels.api as sm\n",
    "logit_model=sm.Logit(yy,XX)\n",
    "result=logit_model.fit()\n",
    "print(result.summary2())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now split the data to see how it perform with the train/test."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression\n",
    "from sklearn import metrics\n",
    "X_train, X_test, y_train, y_test = train_test_split(XX, yy, test_size=0.5, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\sklearn\\utils\\validation.py:1111: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<style>#sk-container-id-1 {color: black;background-color: white;}#sk-container-id-1 pre{padding: 0;}#sk-container-id-1 div.sk-toggleable {background-color: white;}#sk-container-id-1 label.sk-toggleable__label {cursor: pointer;display: block;width: 100%;margin-bottom: 0;padding: 0.3em;box-sizing: border-box;text-align: center;}#sk-container-id-1 label.sk-toggleable__label-arrow:before {content: \"▸\";float: left;margin-right: 0.25em;color: #696969;}#sk-container-id-1 label.sk-toggleable__label-arrow:hover:before {color: black;}#sk-container-id-1 div.sk-estimator:hover label.sk-toggleable__label-arrow:before {color: black;}#sk-container-id-1 div.sk-toggleable__content {max-height: 0;max-width: 0;overflow: hidden;text-align: left;background-color: #f0f8ff;}#sk-container-id-1 div.sk-toggleable__content pre {margin: 0.2em;color: black;border-radius: 0.25em;background-color: #f0f8ff;}#sk-container-id-1 input.sk-toggleable__control:checked~div.sk-toggleable__content {max-height: 200px;max-width: 100%;overflow: auto;}#sk-container-id-1 input.sk-toggleable__control:checked~label.sk-toggleable__label-arrow:before {content: \"▾\";}#sk-container-id-1 div.sk-estimator input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-label input.sk-toggleable__control:checked~label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 input.sk-hidden--visually {border: 0;clip: rect(1px 1px 1px 1px);clip: rect(1px, 1px, 1px, 1px);height: 1px;margin: -1px;overflow: hidden;padding: 0;position: absolute;width: 1px;}#sk-container-id-1 div.sk-estimator {font-family: monospace;background-color: #f0f8ff;border: 1px dotted black;border-radius: 0.25em;box-sizing: border-box;margin-bottom: 0.5em;}#sk-container-id-1 div.sk-estimator:hover {background-color: #d4ebff;}#sk-container-id-1 div.sk-parallel-item::after {content: \"\";width: 100%;border-bottom: 1px solid gray;flex-grow: 1;}#sk-container-id-1 div.sk-label:hover label.sk-toggleable__label {background-color: #d4ebff;}#sk-container-id-1 div.sk-serial::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: 0;}#sk-container-id-1 div.sk-serial {display: flex;flex-direction: column;align-items: center;background-color: white;padding-right: 0.2em;padding-left: 0.2em;position: relative;}#sk-container-id-1 div.sk-item {position: relative;z-index: 1;}#sk-container-id-1 div.sk-parallel {display: flex;align-items: stretch;justify-content: center;background-color: white;position: relative;}#sk-container-id-1 div.sk-item::before, #sk-container-id-1 div.sk-parallel-item::before {content: \"\";position: absolute;border-left: 1px solid gray;box-sizing: border-box;top: 0;bottom: 0;left: 50%;z-index: -1;}#sk-container-id-1 div.sk-parallel-item {display: flex;flex-direction: column;z-index: 1;position: relative;background-color: white;}#sk-container-id-1 div.sk-parallel-item:first-child::after {align-self: flex-end;width: 50%;}#sk-container-id-1 div.sk-parallel-item:last-child::after {align-self: flex-start;width: 50%;}#sk-container-id-1 div.sk-parallel-item:only-child::after {width: 0;}#sk-container-id-1 div.sk-dashed-wrapped {border: 1px dashed gray;margin: 0 0.4em 0.5em 0.4em;box-sizing: border-box;padding-bottom: 0.4em;background-color: white;}#sk-container-id-1 div.sk-label label {font-family: monospace;font-weight: bold;display: inline-block;line-height: 1.2em;}#sk-container-id-1 div.sk-label-container {text-align: center;}#sk-container-id-1 div.sk-container {/* jupyter's `normalize.less` sets `[hidden] { display: none; }` but bootstrap.min.css set `[hidden] { display: none !important; }` so we also need the `!important` here to be able to override the default hidden behavior on the sphinx rendered scikit-learn.org. See: https://github.com/scikit-learn/scikit-learn/issues/21755 */display: inline-block !important;position: relative;}#sk-container-id-1 div.sk-text-repr-fallback {display: none;}</style><div id=\"sk-container-id-1\" class=\"sk-top-container\"><div class=\"sk-text-repr-fallback\"><pre>LogisticRegression()</pre><b>In a Jupyter environment, please rerun this cell to show the HTML representation or trust the notebook. <br />On GitHub, the HTML representation is unable to render, please try loading this page with nbviewer.org.</b></div><div class=\"sk-container\" hidden><div class=\"sk-item\"><div class=\"sk-estimator sk-toggleable\"><input class=\"sk-toggleable__control sk-hidden--visually\" id=\"sk-estimator-id-1\" type=\"checkbox\" checked><label for=\"sk-estimator-id-1\" class=\"sk-toggleable__label sk-toggleable__label-arrow\">LogisticRegression</label><div class=\"sk-toggleable__content\"><pre>LogisticRegression()</pre></div></div></div></div></div>"
      ],
      "text/plain": [
       "LogisticRegression()"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "logreg = LogisticRegression()\n",
    "logreg.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy of logistic regression classifier on test set: 0.99\n"
     ]
    }
   ],
   "source": [
    "y_pred = logreg.predict(X_test)\n",
    "print('Accuracy of logistic regression classifier on test set: {:.2f}'.format(logreg.score(X_test, y_test)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we examine the Confusion Matrix."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[4504    4]\n",
      " [  22    6]]\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import confusion_matrix\n",
    "confusion_matrix = confusion_matrix(y_test, y_pred)\n",
    "print(confusion_matrix)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We now examine some key metrics to see if the model acts properly."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "           0       1.00      1.00      1.00      4508\n",
      "           1       0.60      0.21      0.32        28\n",
      "\n",
      "    accuracy                           0.99      4536\n",
      "   macro avg       0.80      0.61      0.66      4536\n",
      "weighted avg       0.99      0.99      0.99      4536\n",
      "\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import classification_report\n",
    "print(classification_report(y_test, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now we look at the Receiver Operating Characteristic (ROC) curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 5040x1440 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import roc_auc_score\n",
    "from sklearn.metrics import roc_curve\n",
    "logit_roc_auc = roc_auc_score(y_test, logreg.predict(X_test))\n",
    "fpr, tpr, thresholds = roc_curve(y_test, logreg.predict_proba(X_test)[:,1])\n",
    "plt.figure()\n",
    "plt.plot(fpr, tpr, label='Logistic Regression (area = %0.2f)' % logit_roc_auc)\n",
    "plt.plot([0, 1], [0, 1],'r--')\n",
    "plt.xlim([0.0, 1.0])\n",
    "plt.ylim([0.0, 1.05])\n",
    "plt.xlabel('False Positive Rate')\n",
    "plt.ylabel('True Positive Rate')\n",
    "plt.title('Receiver operating characteristic')\n",
    "plt.legend(loc=\"lower right\")\n",
    "plt.savefig('Log_ROC.png')\n",
    "plt.rcParams[\"figure.figsize\"] = (16,9)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Attempting to get the sequence and channel back to a dataframe."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "fd=pd.DataFrame(y_pred)\n",
    "fd2=pd.DataFrame(y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   0\n",
       "0  0\n",
       "1  0\n",
       "2  0\n",
       "3  0\n",
       "4  0"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "fd2.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "FD2=fd2.reset_index()\n",
    "FD2.set_axis(['index','FLAG_ORIG'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "merge=FD2.merge(fd,left_index=True, right_index=True)\n",
    "merge.set_axis(['index','FLAG_ORIG','PREDICT'],axis=1,inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>FLAG_ORIG</th>\n",
       "      <th>PREDICT</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>4</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index  FLAG_ORIG  PREDICT\n",
       "0      0          0        0\n",
       "1      1          0        0\n",
       "2      2          0        1\n",
       "3      3          0        0\n",
       "4      4          0        0"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "merge.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We create a header to spot the different predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "merge['DIFF']=merge['FLAG_ORIG']-merge['PREDICT']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0       False\n",
       "1       False\n",
       "2        True\n",
       "3       False\n",
       "4       False\n",
       "        ...  \n",
       "4531    False\n",
       "4532    False\n",
       "4533    False\n",
       "4534    False\n",
       "4535    False\n",
       "Name: DIFF, Length: 4536, dtype: bool"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "finder=merge['DIFF']!=0\n",
    "finder"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We check what has been predicted different from the test data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index        2\n",
      "FLAG_ORIG    0\n",
      "PREDICT      1\n",
      "DIFF        -1\n",
      "Name: 2, dtype: int64\n",
      "index        386\n",
      "FLAG_ORIG      1\n",
      "PREDICT        0\n",
      "DIFF           1\n",
      "Name: 386, dtype: int64\n",
      "index        490\n",
      "FLAG_ORIG      0\n",
      "PREDICT        1\n",
      "DIFF          -1\n",
      "Name: 490, dtype: int64\n",
      "index        546\n",
      "FLAG_ORIG      1\n",
      "PREDICT        0\n",
      "DIFF           1\n",
      "Name: 546, dtype: int64\n",
      "index        644\n",
      "FLAG_ORIG      1\n",
      "PREDICT        0\n",
      "DIFF           1\n",
      "Name: 644, dtype: int64\n",
      "index        705\n",
      "FLAG_ORIG      1\n",
      "PREDICT        0\n",
      "DIFF           1\n",
      "Name: 705, dtype: int64\n",
      "index        1213\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1213, dtype: int64\n",
      "index        1309\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1309, dtype: int64\n",
      "index        1405\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1405, dtype: int64\n",
      "index        1453\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1453, dtype: int64\n",
      "index        1483\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1483, dtype: int64\n",
      "index        1593\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1593, dtype: int64\n",
      "index        1818\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 1818, dtype: int64\n",
      "index        2221\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 2221, dtype: int64\n",
      "index        2439\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 2439, dtype: int64\n",
      "index        2507\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 2507, dtype: int64\n",
      "index        2715\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 2715, dtype: int64\n",
      "index        3247\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 3247, dtype: int64\n",
      "index        3273\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 3273, dtype: int64\n",
      "index        3319\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 3319, dtype: int64\n",
      "index        3396\n",
      "FLAG_ORIG       0\n",
      "PREDICT         1\n",
      "DIFF           -1\n",
      "Name: 3396, dtype: int64\n",
      "index        3777\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 3777, dtype: int64\n",
      "index        4114\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 4114, dtype: int64\n",
      "index        4124\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 4124, dtype: int64\n",
      "index        4178\n",
      "FLAG_ORIG       1\n",
      "PREDICT         0\n",
      "DIFF            1\n",
      "Name: 4178, dtype: int64\n",
      "index        4393\n",
      "FLAG_ORIG       0\n",
      "PREDICT         1\n",
      "DIFF           -1\n",
      "Name: 4393, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "for i,x in merge.iterrows():\n",
    "    if x['DIFF']!=0:\n",
    "        print(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "X1_train, X1_test, y1_train, y1_test = train_test_split(df1, df1, test_size=0.5, random_state=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "FD3=y1_test.reset_index()\n",
    "newtest=pd.merge(FD3,merge)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>index</th>\n",
       "      <th>SEQ</th>\n",
       "      <th>CHANNEL</th>\n",
       "      <th>DEEP_RMS_x</th>\n",
       "      <th>BAD_CHAN_FLAG</th>\n",
       "      <th>CAP_CUT_ERR</th>\n",
       "      <th>CAP_ERR</th>\n",
       "      <th>CAP_VAL</th>\n",
       "      <th>DEEP_RMS_y</th>\n",
       "      <th>DEVICE</th>\n",
       "      <th>LEAK_ERR</th>\n",
       "      <th>LEAK_VAL</th>\n",
       "      <th>MARG_CHAN_FLAG</th>\n",
       "      <th>SENS</th>\n",
       "      <th>FLAG_ORIG</th>\n",
       "      <th>PREDICT</th>\n",
       "      <th>DIFF</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>247</td>\n",
       "      <td>1</td>\n",
       "      <td>248</td>\n",
       "      <td>2.415203</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>244.630630</td>\n",
       "      <td>8.29</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>31.94</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1972</td>\n",
       "      <td>1</td>\n",
       "      <td>1973</td>\n",
       "      <td>1.955668</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.870605</td>\n",
       "      <td>7.63</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>32.48</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1829</td>\n",
       "      <td>1</td>\n",
       "      <td>1830</td>\n",
       "      <td>0.956624</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>232.227005</td>\n",
       "      <td>4.51</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>22.92</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>154</td>\n",
       "      <td>1</td>\n",
       "      <td>155</td>\n",
       "      <td>1.543170</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>242.414124</td>\n",
       "      <td>5.47</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>30.08</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1278</td>\n",
       "      <td>1</td>\n",
       "      <td>1279</td>\n",
       "      <td>0.922203</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>236.883118</td>\n",
       "      <td>3.54</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>5.0</td>\n",
       "      <td>0</td>\n",
       "      <td>22.35</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   index  SEQ  CHANNEL  DEEP_RMS_x  BAD_CHAN_FLAG  CAP_CUT_ERR  CAP_ERR  \\\n",
       "0    247    1      248    2.415203              0            0        0   \n",
       "1   1972    1     1973    1.955668              0            0        0   \n",
       "2   1829    1     1830    0.956624              0            0        0   \n",
       "3    154    1      155    1.543170              0            0        0   \n",
       "4   1278    1     1279    0.922203              0            0        0   \n",
       "\n",
       "      CAP_VAL  DEEP_RMS_y  DEVICE  LEAK_ERR  LEAK_VAL  MARG_CHAN_FLAG   SENS  \\\n",
       "0  244.630630        8.29       0         0       5.0               0  31.94   \n",
       "1  232.870605        7.63       0         0       5.0               0  32.48   \n",
       "2  232.227005        4.51       0         0       5.0               0  22.92   \n",
       "3  242.414124        5.47       0         0       5.0               0  30.08   \n",
       "4  236.883118        3.54       0         0       5.0               0  22.35   \n",
       "\n",
       "   FLAG_ORIG  PREDICT  DIFF  \n",
       "0          0        0     0  \n",
       "1          0        0     0  \n",
       "2          0        0     0  \n",
       "3          0        0     0  \n",
       "4          0        0     0  "
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "newtest.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index             4393.000000\n",
      "SEQ                  1.000000\n",
      "CHANNEL           4394.000000\n",
      "DEEP_RMS_x           1.446176\n",
      "BAD_CHAN_FLAG        0.000000\n",
      "CAP_CUT_ERR          0.000000\n",
      "CAP_ERR              0.000000\n",
      "CAP_VAL            226.530472\n",
      "DEEP_RMS_y           5.140000\n",
      "DEVICE               0.000000\n",
      "LEAK_ERR             0.000000\n",
      "LEAK_VAL             5.000000\n",
      "MARG_CHAN_FLAG       0.000000\n",
      "SENS                23.550000\n",
      "FLAG_ORIG            0.000000\n",
      "PREDICT              1.000000\n",
      "DIFF                -1.000000\n",
      "Name: 86, dtype: float64\n",
      "index             2272.000000\n",
      "SEQ                  1.000000\n",
      "CHANNEL           2273.000000\n",
      "DEEP_RMS_x           2.109680\n",
      "BAD_CHAN_FLAG        0.000000\n",
      "CAP_CUT_ERR          0.000000\n",
      "CAP_ERR              0.000000\n",
      "CAP_VAL            247.254501\n",
      "DEEP_RMS_y           9.840000\n",
      "DEVICE               0.000000\n",
      "LEAK_ERR             0.000000\n",
      "LEAK_VAL             5.000000\n",
      "MARG_CHAN_FLAG       0.000000\n",
      "SENS                31.780000\n",
      "FLAG_ORIG            1.000000\n",
      "PREDICT              1.000000\n",
      "DIFF                 0.000000\n",
      "Name: 507, dtype: float64\n",
      "index             2866.000000\n",
      "SEQ                  1.000000\n",
      "CHANNEL           2867.000000\n",
      "DEEP_RMS_x           1.387511\n",
      "BAD_CHAN_FLAG        0.000000\n",
      "CAP_CUT_ERR          0.000000\n",
      "CAP_ERR              0.000000\n",
      "CAP_VAL            240.321045\n",
      "DEEP_RMS_y          10.050000\n",
      "DEVICE               0.000000\n",
      "LEAK_ERR             0.000000\n",
      "LEAK_VAL             5.000000\n",
      "MARG_CHAN_FLAG       0.000000\n",
      "SENS                33.050000\n",
      "FLAG_ORIG            1.000000\n",
      "PREDICT              1.000000\n",
      "DIFF                 0.000000\n",
      "Name: 1253, dtype: float64\n"
     ]
    }
   ],
   "source": [
    "for i,x in newtest.iterrows():\n",
    "    if x['PREDICT']>x['BAD_CHAN_FLAG']:\n",
    "        print(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Now for comparison we look at a simple lr , lr L1 norm and lr L2 norm Logistic Regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\sklearn\\utils\\validation.py:1111: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    }
   ],
   "source": [
    "# Standard logistic regression\n",
    "lr = LogisticRegression(solver='liblinear').fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\sklearn\\utils\\validation.py:1111: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    }
   ],
   "source": [
    "from sklearn.linear_model import LogisticRegressionCV\n",
    "# L1 regularized logistic regression\n",
    "lr_l1 = LogisticRegressionCV(Cs=10, cv=4, penalty='l1', solver='liblinear').fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\Users\\sirproc\\Anaconda3\\envs\\data_science\\lib\\site-packages\\sklearn\\utils\\validation.py:1111: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    }
   ],
   "source": [
    "# L2 regularized logistic regression\n",
    "lr_l2 = LogisticRegressionCV(Cs=10, cv=4, penalty='l2', solver='liblinear').fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>lr</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   lr  l1  l2\n",
       "0   0   0   0\n",
       "1   0   0   0\n",
       "2   1   0   0\n",
       "3   0   0   0\n",
       "4   0   0   0"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Predict the class and the probability for each\n",
    "y_pred = list()\n",
    "y_prob = list()\n",
    "coeff_labels = ['lr', 'l1', 'l2']\n",
    "coeff_models = [lr, lr_l1, lr_l2]\n",
    "for lab,mod in zip(coeff_labels, coeff_models):\n",
    "    y_pred.append(pd.Series(mod.predict(X_test), name=lab))\n",
    "    y_prob.append(pd.Series(mod.predict_proba(X_test).max(axis=1), name=lab))\n",
    "y_pred = pd.concat(y_pred, axis=1)\n",
    "y_prob = pd.concat(y_prob, axis=1)\n",
    "\n",
    "y_pred.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>lr</th>\n",
       "      <th>l1</th>\n",
       "      <th>l2</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.996878</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.980332</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.997633</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.985511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.508511</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.783742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.996893</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.979379</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.998146</td>\n",
       "      <td>0.5</td>\n",
       "      <td>0.985073</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "         lr   l1        l2\n",
       "0  0.996878  0.5  0.980332\n",
       "1  0.997633  0.5  0.985511\n",
       "2  0.508511  0.5  0.783742\n",
       "3  0.996893  0.5  0.979379\n",
       "4  0.998146  0.5  0.985073"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_prob.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "ename": "SyntaxError",
     "evalue": "invalid syntax (3685462447.py, line 10)",
     "output_type": "error",
     "traceback": [
      "\u001b[1;36m  Input \u001b[1;32mIn [61]\u001b[1;36m\u001b[0m\n\u001b[1;33m    accuracy = accuracy_score(y_test, y_pred[lab])\u001b[0m\n\u001b[1;37m    ^\u001b[0m\n\u001b[1;31mSyntaxError\u001b[0m\u001b[1;31m:\u001b[0m invalid syntax\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import precision_recall_fscore_support as score\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score, roc_auc_score\n",
    "from sklearn.preprocessing import label_binarize\n",
    "\n",
    "metrics = list()\n",
    "cm = dict()\n",
    "\n",
    "for lab in coeff_labels:\n",
    "    \n",
    "    # Preciision, recall, f-score from the multi-class support function\n",
    "    precision, recall, fscore, _ = score(y_test, y_pred[lab], average='weighted'\n",
    "    \n",
    "    # The usual way to calculate accuracy\n",
    "    accuracy = accuracy_score(y_test, y_pred[lab])\n",
    "    \n",
    "    # ROC-AUC scores can be calculated by binarizing the data\n",
    "#   auc = roc_auc_score(label_binarize(y_test, classes=[0,1,2,3,4,5]),\n",
    "#       label_binarize(y_pred[lab], classes=[0,1,2,3,4,5]),\n",
    "#       average='weighted')\n",
    "    \n",
    "    # Last, the confusion matrix\n",
    "    cm[lab] = confusion_matrix(y_test, y_pred[lab])\n",
    "\n",
    "    metrics.append(pd.Series({'precision':precision, 'recall':recall,\n",
    "                                'fscore':fscore, 'accuracy':accuracy},\n",
    "#                                   auc':auc},\n",
    "                                name=lab))\n",
    "metrics = pd.concat(metrics, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Checking the results.."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<module 'sklearn.metrics' from 'c:\\\\Users\\\\sirproc\\\\Anaconda3\\\\envs\\\\data_science\\\\lib\\\\site-packages\\\\sklearn\\\\metrics\\\\__init__.py'>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "metrics"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Lastly, the three confusion matrices:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'cm' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32mz:\\Dan\\Code\\Seismic\\Bad_Channels_Logistic_Regression.ipynb Cell 87'\u001b[0m in \u001b[0;36m<cell line: 7>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000058?line=4'>5</a>\u001b[0m axList[\u001b[39m-\u001b[39m\u001b[39m1\u001b[39m]\u001b[39m.\u001b[39maxis(\u001b[39m'\u001b[39m\u001b[39moff\u001b[39m\u001b[39m'\u001b[39m)\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000058?line=6'>7</a>\u001b[0m \u001b[39mfor\u001b[39;00m ax,lab \u001b[39min\u001b[39;00m \u001b[39mzip\u001b[39m(axList[:\u001b[39m-\u001b[39m\u001b[39m1\u001b[39m], coeff_labels):\n\u001b[1;32m----> <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000058?line=7'>8</a>\u001b[0m     sns\u001b[39m.\u001b[39mheatmap(cm[lab], ax\u001b[39m=\u001b[39max, annot\u001b[39m=\u001b[39m\u001b[39mTrue\u001b[39;00m, fmt\u001b[39m=\u001b[39m\u001b[39m'\u001b[39m\u001b[39md\u001b[39m\u001b[39m'\u001b[39m,cmap\u001b[39m=\u001b[39mcmap1);\n\u001b[0;32m      <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000058?line=8'>9</a>\u001b[0m     ax\u001b[39m.\u001b[39mset(title\u001b[39m=\u001b[39mlab);\n\u001b[0;32m     <a href='vscode-notebook-cell:/z%3A/Dan/Code/Seismic/Bad_Channels_Logistic_Regression.ipynb#ch0000058?line=10'>11</a>\u001b[0m plt\u001b[39m.\u001b[39mtight_layout()\n",
      "\u001b[1;31mNameError\u001b[0m: name 'cm' is not defined"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 864x720 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, axList = plt.subplots(nrows=2, ncols=2)\n",
    "axList = axList.flatten()\n",
    "fig.set_size_inches(12, 10)\n",
    "\n",
    "axList[-1].axis('off')\n",
    "\n",
    "for ax,lab in zip(axList[:-1], coeff_labels):\n",
    "    sns.heatmap(cm[lab], ax=ax, annot=True, fmt='d',cmap=cmap1);\n",
    "    ax.set(title=lab);\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Conclsions**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "37e0769f8301bc1fd40520bc2e83c2f2ee17339fb59258be934dee8f147fee84"
  },
  "kernelspec": {
   "display_name": "Python 3.9.12 ('data_science')",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.12"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}