Programming #1 MNIST_hiddenlayer_backprop.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np #load library first always\n",
    "import scipy.linalg as la #for eigenvalues\n",
    "import matplotlib.pyplot as plt #to be able to plot \n",
    "from numpy import genfromtxt #import data\n",
    "import pandas as pd  #for finding max and min values, confusion matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#######################################\n",
    "#\n",
    "# Importing Data\n",
    "# I added 'import from numpy genfromtxt'.  \n",
    "# I made found mnist_test and mnist_train on https://github.com/pjreddie/mnist-csv-png. \n",
    "#\n",
    "#######################################\n",
    "\n",
    "\n",
    "my_test_data = genfromtxt('/Users/vickyhaney/Desktop/mnist_test.csv', delimiter=',') \n",
    "#sizetest = len(my_test_data)\n",
    "#print(my_test_dat.shape)\n",
    "my_train_data = genfromtxt('/Users/vickyhaney/Desktop/mnist_train.csv', delimiter=',') \n",
    "#sizetrain = len(my_train_data) #takes a long time\n",
    "#print(size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10000, 785)\n",
      "(60000, 785)\n"
     ]
    }
   ],
   "source": [
    "print(my_test_data.shape)\n",
    "print(my_train_data.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 1)\n",
      "(10000, 1)\n"
     ]
    }
   ],
   "source": [
    "#let 'index' be the 1x10000 vector from \n",
    "#my_test_data which represents my 0-9 integers\n",
    "##################################\n",
    "#\n",
    "#  index labels for training data\n",
    "#\n",
    "##################################\n",
    "\n",
    "index_train = my_train_data[:,0:1]\n",
    "print(index_train.shape)\n",
    "index_train = index_train.astype(int)  #convert to int from float\n",
    "\n",
    "\n",
    "\n",
    "#let 'index' be the 1x10000 vector from \n",
    "#my_test_data which represents my 0-9 integers\n",
    "##################################\n",
    "#\n",
    "#  index labels for test data\n",
    "#\n",
    "##################################\n",
    "\n",
    "index_test = my_test_data[:,0:1]\n",
    "print(index_test.shape)\n",
    "index_test = index_test.astype(int)  #convert to int from float\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 785)\n",
      "(785, 60000)\n",
      "(10000, 785)\n"
     ]
    }
   ],
   "source": [
    "##################################\n",
    "#\n",
    "#  Features_Training table placing bias \n",
    "#  in first row instead of integers\n",
    "#\n",
    "##################################\n",
    "#Let F(features) equal the 60000 X 784 features scaled \n",
    "#by 255 and replace the 1st column with 1s\n",
    "\n",
    "f_train = my_train_data/255\n",
    "f_train[:,0] = np.ones((60000,)) #making the first row all 1s \n",
    "print(f_train.shape)    #correct size :) 60000 x785 columns\n",
    "\n",
    "F_train = np.transpose(f_train)    #785 x 600000\n",
    "print(F_train.shape)\n",
    "\n",
    "#print(F_train)\n",
    "\n",
    "##################################\n",
    "#\n",
    "#  Features_test data table placing bias in first row instead of integers\n",
    "#\n",
    "##################################\n",
    "#Let F(features) equal the 10000 X 784 features scaled by 255 and replace the 1st column with 1s\n",
    "f_test = my_test_data/255\n",
    "f_test[:,0] = np.ones((10000,)) #making the first row all 1 \n",
    "print(f_test.shape)    #correct size :) 10000 x785 columns\n",
    "\n",
    "F_test = np.transpose(f_test)   #785 x 10000\n",
    "#print(F_test.shape)\n",
    "#print(F_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(60000, 1)\n",
      "(10, 60000)\n",
      "(10000, 1)\n",
      "(10, 10000)\n"
     ]
    }
   ],
   "source": [
    "###############################\n",
    "#\n",
    "# This is my Truth_train matrix. I used my index_train from above. \n",
    "#  It will never change. \n",
    "# 1-hot encoded\n",
    "# https://www.kite.com/python/answers/how-to-do-one-hot-encoding-with-numpy-in-python\n",
    "#\n",
    "###############################\n",
    "\n",
    "\n",
    "\n",
    "print(index_train.shape)\n",
    "\n",
    "shape = (np.transpose(index_train).size, np.transpose(index_train).max()+1)            \n",
    "                                                \n",
    "t_train = np.zeros(shape)                    #creates a matrix of zero with shape I need\n",
    "rows = np.arange(np.transpose(index_train).size)\n",
    "t_train[rows, np.transpose(index_train)] = 1\n",
    "\n",
    "#print(t_train.shape)\n",
    "T_train = np.transpose(t_train)\n",
    "#print(T_train)\n",
    "#print(T_train.shape)\n",
    "\n",
    "\n",
    "###############################\n",
    "#\n",
    "# This is my Truth_test matrix. I used my index_test from above. \n",
    "# It will never change. \n",
    "# 1-hot encoded\n",
    "# #https://www.kite.com/python/answers/how-to-do-one-hot-encoding-with-numpy-in-python\n",
    "#\n",
    "###############################\n",
    "\n",
    "\n",
    "print(index_test.shape)\n",
    "\n",
    "shape1 = (np.transpose(index_test).size, np.transpose(index_test).max()+1)            \n",
    "                                                \n",
    "t_test = np.zeros(shape1)                     #creates a matrix of zero with shape I need\n",
    "rows1 = np.arange(np.transpose(index_test).size)\n",
    "t_test[rows1, np.transpose(index_test)] = 1\n",
    "\n",
    "#print(t_train.shape)\n",
    "T_test = np.transpose(t_test)\n",
    "#print(T_test)\n",
    "#print(T_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "###################################\n",
    "#\n",
    "# Thresholded my Target matrix to 0.9's and 0.1's\n",
    "# due to Sigmoid function\n",
    "#\n",
    "##################################\n",
    "\n",
    "def threshold(T):\n",
    "    #T[T >= 1] = 0.9   different way of doing this\n",
    "    #T[T < 1] = 0.1    different way of doing this\n",
    "    T = np.where(T == 1, 0.9, T)         \n",
    "    T = np.where(T <= 0, 0.1, T)\n",
    "    return T\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 60000)\n",
      "(10, 10000)\n"
     ]
    }
   ],
   "source": [
    "###################################\n",
    "#\n",
    "# These are more non-changing Target_train\n",
    "# and Target_test matrices of 0.9's and 0.1's\n",
    "# due to Sigmoid function\n",
    "#\n",
    "##################################\n",
    "\n",
    "T_train = threshold(T_train)\n",
    "#print(T_train)\n",
    "print(T_train.shape)\n",
    "#print(np.min(T_train))\n",
    "#print(np.max(T_train))\n",
    "\n",
    "\n",
    "T_test = threshold(T_test)\n",
    "#print(T_test)\n",
    "print(T_test.shape)\n",
    "#print(np.min(T_test))\n",
    "#print(np.max(T_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(100, 785)\n"
     ]
    }
   ],
   "source": [
    "#####################\n",
    "#\n",
    "# Initializing my weights for \n",
    "# input to hidden: w_hidden \\in (n) x 78\n",
    "#\n",
    "# len(F_Train) is size of features (785) including bias\n",
    "# Write a command to create a random matrix of size 4x4 whose elements are \n",
    "# uniformly distributed in the interval (a,b), \n",
    "# (b - a) * np.random.random_sample((m, n)) + a\n",
    "#\n",
    "######################\n",
    "n = 100 # Hidden Nodes: 20, 50,100\n",
    "\n",
    "\n",
    "w_hidden = (0.05-(-0.05)) * np.random.random_sample((n, len(F_train))) + (-0.05)\n",
    "print(w_hidden.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(10, 101)\n"
     ]
    }
   ],
   "source": [
    "#####################\n",
    "#\n",
    "# Initializing my weights for \n",
    "# hidden to output: w_output \\in (10) x (n+1) (bias on hidden node)\n",
    "#\n",
    "# Write a command to create a random matrix of size 4x4 whose elements are \n",
    "# uniformly distributed in the interval (a,b), \n",
    "# (b - a) * np.random.random_sample((m, n)) + a\n",
    "#\n",
    "######################\n",
    "\n",
    "#take into account hidden node bias\n",
    "\n",
    "w_outer = (0.05-(-0.05)) * np.random.random_sample((10,n+1)) + (-0.05)\n",
    "print(w_outer.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#################################\n",
    "#\n",
    "#initialize previous delta_weigths as zeros\n",
    "#\n",
    "##################################\n",
    "previous_d_hidden = np.zeros((n,785))\n",
    "previous_d_outer = np.zeros((10,n+1)) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "#####################################\n",
    "#\n",
    "# This is my activation function for my hidden layer\n",
    "# takes our feature matrix multiplied\n",
    "# by my weight matrix, then activitates \n",
    "# using our sigmoid function sigma=1/(1+e^(-x))\n",
    "#\n",
    "######################################\n",
    "\n",
    "\n",
    "\n",
    "def activation(A,B):\n",
    "    y = np.dot(A, B)  \n",
    "    return y\n",
    "\n",
    "\n",
    "# define sigmoid function \n",
    "\n",
    "def sigmoid(y):\n",
    "    s = 1/(1 + np.exp(-y))\n",
    "    return(s)\n",
    "\n",
    "\n",
    "#############################################\n",
    "#\n",
    "# Big Delta weight updates for Backward phase\n",
    "# since we only have one output here, \n",
    "# S_k, k= outer\n",
    "# S_j, j =hidden\n",
    "#\n",
    "###############################################\n",
    "\n",
    "def Output_weight_updates(o, T): #o will need to be indexed for more than 1 output\n",
    "    delta_outer = o * (1 - o) * (T - o)\n",
    "    return delta_outer\n",
    "\n",
    "def Hidden_weight_updates(h,w,d):   #this will be w_outer\n",
    "    delta_hidden = h * (1 - h) * (activation(w,d))   \n",
    "    return delta_hidden\n",
    "\n",
    "\n",
    "#########################################\n",
    "#\n",
    "# This is my delta weights function used for \n",
    "# delta hidden and delta outer\n",
    "#\n",
    "#########################################\n",
    "\n",
    "eta = 0.1\n",
    "alpha = 0.9\n",
    "\n",
    "def delta_weights(eta,alpha,d,h,p):   #eta,alpha(momentum),h = hidden_array, d= D_outer/D_hidden, p=previous weight\n",
    "    delta_w = (eta * np.outer(d , h)) + (alpha * p)\n",
    "    return delta_w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#########################################\n",
    "#\n",
    "# This is my accuracy function \n",
    "#\n",
    "#########################################\n",
    "\n",
    "def accuracy(w_h,w_o,F,T,m):                          #(w_hidden,w_outer,F_train,T_train,60000)\n",
    "    Z = np.zeros((1, m))\n",
    "    for column in range(len(F[0])):                  # for column in range(len(F_train[0])) \n",
    "        Y = activation(w_h,F[:,column:column+1])     # Y = activation(w_hidden,F_train) \n",
    "        H = sigmoid(Y)                               # 20 x 1\n",
    "        H = np.append(1, H)                          #add bias but makes it a row\n",
    "        H = np.reshape(H, (n+1,1)) #reshape to column vector and for some reason it loses the dimension 1    \n",
    "        X = activation(w_o,H)      # becoming 1x2\n",
    "        O_array = sigmoid(X)                #output array will stay 1x2\n",
    "        Z[:,column:column+1] = O_array.argmax() - T[:,column:column+1].argmax()  #Outputs_predicted - T   # Y - trut    \n",
    "    r = np.nonzero(np.any(Z != 0, axis=0))[0]     #find all columns that are nonzero\n",
    "    c = len(r)                                    #find all columns that are nonzero\n",
    "    Acc = 1 - c/m                                 # 1 - number of nonzero columns/size of samples\n",
    "    return Acc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0\n",
      "1\n",
      "2\n",
      "3\n",
      "4\n",
      "5\n",
      "6\n",
      "7\n",
      "8\n",
      "9\n",
      "10\n",
      "11\n",
      "12\n",
      "13\n",
      "14\n",
      "15\n",
      "16\n",
      "17\n",
      "18\n",
      "19\n",
      "20\n",
      "21\n",
      "22\n",
      "23\n",
      "24\n",
      "25\n",
      "26\n",
      "27\n",
      "28\n",
      "29\n",
      "30\n",
      "31\n",
      "32\n",
      "33\n",
      "34\n",
      "35\n",
      "36\n",
      "37\n",
      "38\n",
      "39\n",
      "40\n",
      "41\n",
      "42\n",
      "43\n",
      "44\n",
      "45\n",
      "46\n",
      "47\n",
      "48\n",
      "49\n"
     ]
    }
   ],
   "source": [
    "#################################################\n",
    "#\n",
    "# NEURAL NETWORK ALGORITHM\n",
    "# Depending on which experiment, \n",
    "# plug in F_train[:,i:i+1], F_train_half[:,i:i+1]), F_train_quarter[:,i:i+1])\n",
    "# T_train[:,i:i+1], T_train_half[:,i:i+1]), T_train_quarter[:,i:i+1])\n",
    "# where indicated\n",
    "#\n",
    "##################################################\n",
    "\n",
    "\n",
    "epo = 50 #50\n",
    "\n",
    "eta = 0.1\n",
    "alpha = 0.9  #0, 0.25, 0.5\n",
    "\n",
    "###########################\n",
    "#For accuracy: initialize O_node_array and H_node_array\n",
    "###########################\n",
    "accuracy_training = np.zeros(epo) #number of epochs\n",
    "accuracy_test = np.zeros(epo) #number of epochs\n",
    "\n",
    "\n",
    "for epoch in range(epo):\n",
    "    print(epoch)\n",
    "    idx = np.random.permutation(60000)  #60000, 30000, 15000 \n",
    "    for i in idx:\n",
    "\n",
    "################################\n",
    "#\n",
    "# Making hidden_arrays\n",
    "# Begin Forward prop\n",
    "#\n",
    "################################\n",
    "\n",
    "        y = activation(w_hidden,F_train[:,i:i+1])   #F_train_quarter[:,i:i+1],  F_train_half[:,i:i+1])\n",
    "        hidden = np.transpose(sigmoid(y))   #20x 1, no extra '1'\n",
    "        hidden_array = np.append(1, hidden)\n",
    "        hidden_array = np.reshape(hidden_array, (1, n+1)) #for some reason it loses the dimension 1\n",
    "\n",
    "            \n",
    "################################\n",
    "#\n",
    "# Creating output \n",
    "# using hidden array and w_outer\n",
    "#\n",
    "#################################    \n",
    "    \n",
    "        o_array = activation(w_outer,np.transpose(hidden_array))\n",
    "        output_array = sigmoid(o_array)\n",
    "        \n",
    "################################\n",
    "#\n",
    "# Ends forward phase\n",
    "# Begin Backprop\n",
    "#            \n",
    "#################################\n",
    "    \n",
    "################################\n",
    "#\n",
    "# Obtain Big Delta_outer_error\n",
    "#\n",
    "#################################\n",
    "     \n",
    "        D_outer_array = Output_weight_updates(output_array, T_train[:,i:i+1])   #T_train_quarter[:,i:i+1]) \n",
    "        \n",
    "    \n",
    "################################\n",
    "#\n",
    "# Obtain Big Delta_hidden_error\n",
    "# needs Delta_outer to calc\n",
    "#\n",
    "#################################\n",
    "        \n",
    "        w_outer_del = np.delete(w_outer, 0, 1) #deleted first row that deals with h_0\n",
    "        D_hidden = Hidden_weight_updates(np.transpose(hidden),np.transpose(w_outer_del),D_outer_array)\n",
    "        #this provides a 21 x 1 vector.  I don't care about first one. it has to do with hidden_node bian\n",
    "\n",
    "    \n",
    "################################\n",
    "#\n",
    "# Delta weights from output to hidden\n",
    "#\n",
    "#################################\n",
    "     \n",
    "        d_w_outer = delta_weights(eta,alpha,D_outer_array,hidden_array,previous_d_outer)    \n",
    "                 \n",
    "        w_outer += d_w_outer #over write our w_outer, change to \n",
    "        previous_d_outer = d_w_outer\n",
    "\n",
    "                        \n",
    "################################\n",
    "#\n",
    "# Delta weights from hidden to outer\n",
    "#\n",
    "#################################  \n",
    "\n",
    "\n",
    "        d_w_hidden = delta_weights(eta,alpha,D_hidden,F_train[:,i:i+1],previous_d_hidden)  #F_train_quarter[:,i:i+1] \n",
    "\n",
    "        w_hidden +=  d_w_hidden #over write our w_outer \n",
    "        previous_d_hidden = d_w_hidden #rewrite previous w_hidden                 \n",
    "\n",
    "#######################################\n",
    "#\n",
    "# Accuracy functions\n",
    "#\n",
    "#######################################\n",
    "\n",
    "    accuracy_training[epoch] = accuracy(w_hidden,w_outer,F_train,T_train,15000) #F_train_quarter, T_train_quarter, 30000, 60000\n",
    "\n",
    "    accuracy_test[epoch] = accuracy(w_hidden,w_outer,F_test,T_test,10000)   #(w_hidden,w_outer,F_test,T_test,10000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(5,)\n"
     ]
    }
   ],
   "source": [
    "print(accuracy_training.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b2e147670>"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAdoAAAEICAYAAAD4EjWLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAABgzklEQVR4nO2dd3wcxfXAv0+9W7YkN7l3yx0MhlBs6o/ee28h9ITeQolpJoFAICSQUA0hFNNM6L0aYxk33G25SrItS1bv0vz+mD37JN1JJ+lOd5Le9/PZz93OzM6+2dvbtzPz5j0xxqAoiqIoSmAIC7YAiqIoitKVUUWrKIqiKAFEFa2iKIqiBBBVtIqiKIoSQFTRKoqiKEoAUUWrKIqiKAGkwxStiHwkIhf6u2wwEZGNInJ4sOXoTojIaBFZLCIlInJdsOXpyojIEBExIhLhz7JKcBCR+0Vkp4hsC7Ys3hCRF0Xk/mDL4Q0RmSEiW1t7XLOKVkRK3bZ6Ealw2z+3NScyxhxtjHnJ32VDFX/dMK182F3klD2zvecNYW4BvjLGJBpjngjkiYKpPERksogsFJFy53NyM2WvEZFMEakSkRc7TsrQQ0QOE5FVznX7SkQGN1P2NyLys/PStlREDuxIWVtCRI4Vke9FpFBEtonIsyKS6JYfLSLPi0ixk39DM3UNAm4EMowxfTtC/kDj9v8sbbSF3POvWUVrjElwbcBm4Hi3tP+4yulbbMhwIVAAXNCRJ+3g338wsLwtB3aW+1REooD3gFeAnsBLwHtOuidygPuB5ztGwtBERFKBt4G7gF5AJvC6l7K9gPeBvwDJwJ+B90WkZ4cI6xs9sL9rf2AskI6V18W9wEjsf+IQ4BYROcpLXYOAfGPMjtYK0Qn+N8nuusoY4/E3DyrGGJ82YCNwuPN9BrAVuBXYBryMfSD8D8gDdjnfB7gd/zVwmfP9IuB74BGn7Abg6DaWHQp8C5QAnwNPAa94aYMvMt4H/ODU9ymQ6pZ/PrAJyAfudL8mjc5zOVADVAOlwPtOen/gLef8G4Dr3I7ZF/tgKAa2A3910jcDxqmnFNjfS9sGA/XAqUAt0NctLxy4A1jvtGshMNDJGwd8hlXQ24E7nPQXgfvd6pgBbG10P9wKLAWqgAjgNrdzrABObiTjb4GVbvl7ATcDbzUq9wTwNw9t/BKoAyqdazEK+zCa7VzTTcAfgTC3e+cH4DHnN7vfQ51hbnLnA28Avbxde2C4I0c+sBP4D/aP7vN/yYf/2pFANiBuaZuBo1o47n7gxVae61hgkXPfbQHudcsb4rQ/wu3/8RDws1P+Pbdr5Sp7oSPrTuDORvf3PKAQyAX+DkT5+bpdDvzoth8PVABjPJQ9DljeKG0NcKmP5/qaZp4VgdiAU4Blbvs5wJFu+/cBr3k47nDnOtQ79/GLTvoJ2JfWQqc9Y92O2Uij/7eHesew59mxGjjDl/vKyT8Q+NE59xbgIif9Rewz/APnus4Hhnu5Hg3uTw/5LwJPOzKWAN8Ag93yfwMsAIqcz9+45fUCXnCu8S7gXSd9Blb33QjscO7li1v87VrxI2+koaKtBR4GooFYIAX7kI8DEoE3XcK53ZjuyrMG++ANB650GiRtKDsPq4SjnB+vGO+K1hcZ12Mf4LHO/iwnLwN7kx7stPmvzjVoomjdfmR3RRWGVXB3O7IOA7KA/3Nrx/nO9wRgP19uJrf67wJ+dr4vA250y7vZSRsNCDDJuRaJzo1yIxDj7E/zIv8MmiraxcBAINZJOx37MhEGnAmUAf3c8rKBfRwZRmBfDvo55ZKdchHYG3jvZh5wl7ntz8Y+8BOda7X7YYm9d2qBa516Yz3U93vgJ2CA87s+A/zX27V35D7CKZuGfcl7vJnfZSn2YeJp+4eXY64HPmqU9j/339TLcW1RtDOACc5vNhH7snWSp/Y71z4bGI9VYm/h/Nfcyv4b+9+ZhH1Aj3Xy9wb2c36HIdgXrj80I5e3a1YI3OblmL8B/2yU9itwqoeyxwErGqWtBR7z8bp9jZdnhYeyg1pozzk+nvNxHEWK7TQYoI9b/mm4KWIPv7P7/3cU9n93BBCJnZJZh/Pyg4f/d6P64rEK8mLnN52CfbnK8OG+GoxVfGc7504BJrs9d/KxL2YR2BfZJi8P3v6fjfJfdM7jemb/DfjeyeuFVaDnO+c529lPcfI/wI6G9HRknO7WrlpgppN+DFAO9Gz2t2vFH3IjDRVtNRDTTPnJwK5GN6a78lznlhfnXLC+rSmLvYFrgTi3/Ffwomh9lPGPbvtXAR873+92/8GdG60a3xXtNGBzozK3Ay84378F/kSjt+KWbia3cmtxHlxOvUvc8lYDJ3o45mxgkY/yz6Cpor2kBZkWu84LfAL83ku5j4DfOt+bPAAblXW/N8Kd3yDDLf93wNdu987mFmRcCRzmtt8P+2LnUgjNXnvgJG/XsK0b9qXptUZp/6FRr8DDca1WtB7qeBxH2TRuP42UCfbls9r5HVxl3UeIfgbO8nKePwDv+Pm6PUcjZYftcV7koWwKVsm5HvYXYnt8z/h4rq/x8qwIxIZViLuAUc7+QOd6xzQqs9HL8Y3/v3cBb7jth2FfomY4+83+v7Ev0t81SnsGuMeH++p2b7899rnzrNv+McAqL2Vd91xho22sW13uz+wE7IjYQKyC/blRffOwz4x+zr3QRHk617GChi/fO3A6Rt629lgd5xljKl07IhInIs+IyCYRKcYqjmQRCfdy/G7LN2NMufM1oZVl+wMFbmlg37I84qOM7hZ55W4y9Xev2xhThn3z8pXBQH/HsKFQRAqxw7l9nPxLsW+Zq0RkgYgc52vFInIAdgj9NSfpVWCCmwHNQOzbd2O8pftKg2stIhc4FsGu9o0HUn0410vAec7387BTEb6Qin1IbnJL24Sdy/IoowcGA++4ybwS+2fs46mwiPQRkddEJNu5h15hTxv9RSmQ1CgtCft27ldEZJpjNJQnIkXAFTTfHvfruQl7/d3Le/z/iMgoEfmfY7RTDDzYwnnags/XzRiTD5wI3IDtbR2FnXpqjUWpt2eFXxGR/bD/6dOMMWuc5FLn0729rblH+uP2vzHG1GN/W1//O4OBaY2eZ+diO0At3VctPXdae11TjTHJbttKT20wxpRih7n706j9Dq5nx0CsXtnl5Xz5xpja1sjYHkVrGu3fiB2anGaMScJ218EOEwaKXKCXiMS5pQ1spnx7ZMx1r9s5Z0oz5Rtfny3AhkY3RKIx5hgAY8xaY8zZQG/skPwcEYn3UI8nLnTasNgx3Z/vlu4693APx23BDmF7ogw7euDCk6Xibtkc685/A9dgh1+SscN2rmvrTQaAd4GJIjIe26P9j5dyjdmJ7X0OdksbhH0zbyKjF7Zg5/zdf5cYY0y2l2MfdNInOPfQeTRz/4jIcg9Wka7taS+HLcdeD/d6J9JGI7AWeBWYi52z74Gd02ru/+D+/xqEvf47fTjPP4FVwEjnut3R3HmauWalInKHl8OWY4esXXXEY+85j9fNGPONMWYfY0wvbA9nDLYX7ldEZFAL7fG6gkNEpmB/n0uMMV+4yb4L+0ya5FZ8Er7fIzm4/W+ce20gvv93tgDfNPrfJBhjrnTym7uvmnsW+Bv3Z3YCdsg4h0btd3A9O7Zg9Uqyv4Tw5zraRGyXutCx6LvHj3V7xBizCWtAdK+IRInI/sDxAZJxDnCciBzoWH/OpPnrt52GSuxnoEREbhWRWBEJF5HxIrIPgIicJyJpzptloXNMPdbIpx4vClFEYoAzsIYgk922a4FzHIvBZ4H7RGSkWCaKSAp23q+fiPxB7FKBRBGZ5lS9GDhGRHqJSF/sUF9zuF4K8hy5Lsb2aF08C9wkIns7MoxwlDPOyMgc7J/zZ2PM5hbOhXNcHdZ46QFH9sHYHsorvhzv8LRz/GBH7jQROdHJ83TtE7G9iSIRScfOfzcn4zjT0CLSfbvCy2FfY3vV1zm/yzVO+peeCotIhHMfhAPhIhIjbpaiYpdAzPByrkTs23uliOwLnNNce4DzRCTDedGcCcxxfoeWSMTaT5SKyBisrYVXmrlmCcaYB70c9g4wXkROda7H3cBSY8wqT4VFZIqIRIpIEtbOY4sx5hMnz7V0ZIgPbWsWY8zmFtrj8cXSefH8GLjWGPO+hyKzgT+KSE/nmv4WO1zqC28Ax4pdDhWJ7YRUYQ2UfOF/wCgROd+5hpEiso+IjHXym7uv/gMcLiJnOPduijSzfK2dHOP2zL4P+MkYswX40JH/HEeGM7FTIf8zxuRip7P+4VzbSBE52PspWsafivZxrFHATqxxycd+rLs5zsVag+Zj56hex94wnnicNspojFkOXI1VBrnY+ZLmhpmeAzLEDqu86zyMjsMqwQ2ODM9irWbBDl0tF5FS7KT9WcaYCmdY/AHgB6eu/Rqd5yTsy8NsY8w214Zd6hHh1PtX7B/rU+zD7jmsgUMJdl7neOxwzVrsMgGww7dLsHM1n+JlmYTb9VkBPIqd59iONYT4wS3/Tacdr2KHt97Fvl26eMk5xtdhYxfXYnvfWVjr9Fdp3TKXv2HfvD8VkRLsfTHNkdnTtf8T1lq6CGsw8XYr5W0RY0w19ne9APvSdQnWkKQaQETuEJGP3A75I/YeuA3bw65w0hCRgdjrvczL6a4CZjptvxt7nzTHy9iH+TasAZ2vTkNuwj5sS7AjH35fgmGMycMaOz6A/X9OA85y5YvI041GEW7B/g+3YOflTnbLG4gdSnTv4XU0N2IN7p5z6/2691jvwQ7BbsJa1P7FGOPTM80Ysxp7rzyJvQbHY5dvVvt4fAnWOv4sbO9wG3uMY6GZ+8p5kT7GaV8B9qXevWfeWgobjRC4ryd+FXudCrAGeec5MuRjn8c3YnXHLcBxxhjX6Mz52NGaVdg52D+0Q77dlrtdBhF5HTt5HvAeteI/xC6oX4U1iCsOtjxdBRE5DxhnjLndD3V9jTU0fLbdgoU4IvJHrB3KM8GWRWkbYp23bDXG/DHYsoT6QuQWcYZeC7C9xCOxBg6zgiqU0ipEJAw75PuaKln/YoxpzTC64mCMCVk3gErno9MrWqyRzttYw6StwJXGmEXBFUnxFbEGK9uxw1/evNooiqJ0Wrrc0LGiKIqihBIaJk9RFEVRAkhXGDoOCKmpqWbIkCHBFkNRFKVTsXDhwp3GmLRgyxFKqKL1wpAhQ8jMzAy2GIqiKJ0KEWnscanbo0PHiqIoihJAVNEqiqIoSgAJOUUrIkeJyGoRWScit3nIHywiX4jIUhH5WkQGuOUNEpFPRWSliKxwuU8TkaEiMt+p83XxHkBbURRFUfxKSM3Rio2i8xTWLeBWYIGIzHXc+7l4BOtu8CURORQbiPp8J2828IAx5jOxDqTrnfSHsSGaXnNcsF2KdXLeKmpqati6dSuVlZUtF+7kxMTEMGDAACIjI4MtiqIoSqcmpBQtNtjvOmNMFoCIvIb19OSuaDOwXoQAvsL6zEVEMrAxAj+D3SGRXFEpDmWPU+uXgHtpg6LdunUriYmJDBkyBJFABiUKLsYY8vPz2bp1K0OHDg22OIqiKJ2aUBs6TqdhDMStNIyPCNbR/SnO95OBRLGRaEZhnUu/LSKLROQvTg85BSh0ix/oqU4ARORyEckUkcy8vLwm+ZWVlaSkpHRpJQsgIqSkpHSLnruiKEqgCTVF6ws3AdNFZBEwHRtdow7bOz/Iyd8HG9rsotZUbIz5lzFmqjFmalqa52VgXV3Juugu7VQURQk0oaZos2kYWHoAjcJUGWNyjDGnGGOmAHc6aYXYnupiY0yW03t9FxvOLB9IdovP2aRORVEUBaqrQb3y+p9QU7QLgJGOlXAUNtbhXPcCIpLqRHsBuJ09sUcXYBWqqyt6KLDCWGfOXwGnOekXAu8FsA0BIz8/n8mTJzN58mT69u1Lenr67v3q6pbDSH799df8+KOvcZ0VRelu3HMPzJgBVd4ieittIqQUrdMTvQb4BFgJvGGMWS4iM0XkBKfYDGC1iKwB+mCDPOMEVr8J+EJElgGCDTANcCtwg4isw87ZPtdBTfIrKSkpLF68mMWLF3PFFVdw/fXX796Pimp5xZIqWkVRvLFqFTz6KAwbBtHRLZdXfCfUrI4xxnwIfNgo7W6373OAOV6O/QyY6CE9C2vR3OVYuHAhN9xwA6WlpaSmpvLiiy/Sr18/nnjiCZ5++mkiIiLIyMhg1qxZPP3004SHh/PKK6/w5JNPctBBBwVbfEVRQoSnnoK4OHj44WBL0vUIOUXbWVi79g+Uli72a50JCZMZOfJxn8sbY7j22mt57733SEtL4/XXX+fOO+/k+eefZ9asWWzYsIHo6GgKCwtJTk7miiuuICEhgZtuusmvciuK0vl5/HH43e+gd+9gS9L1UEXbiamqquLXX3/liCOOAKCuro5+/foBMHHiRM4991xOOukkTjrppCBKqShKKFNWZudke/WC8eODLU3XRBVtG2lNzzNQGGMYN24c8+bNa5L3wQcf8O233/L+++/zwAMPsGzZsiBIqCiKv6ivh7//Hc44A/r29V+9M2fCCy/A6tXQs6f/6lX2EFLGUErriI6OJi8vb7eirampYfny5dTX17NlyxYOOeQQHn74YYqKiigtLSUxMZGSkpIgS60oSlt49134/e/h1lv9V+fKlfDXv8Jxx6mSDSSqaDsxYWFhzJkzh1tvvZVJkyYxefJkfvzxR+rq6jjvvPOYMGECU6ZM4brrriM5OZnjjz+ed955h8mTJ/Pdd98FW3xFUVqBMytEVpZ/6jMGrrkGEhJg1iz/1Kl4RoeOOyn33nvv7u/ffvttk/zvv/++SdqoUaNYunRpIMVSFCVA7L+/7dE+84x1LOHDir5meeMN+PJLa22sBlCBRXu0iqJ0edassUOknZXZsyE7Gw4+GCorITOz/XV+8w1MmWItjZXAoopWUZQuz8cfw403wq+/BluS1rN2LVx0EfzjH3DEEbB0Key3X9vq2rYNPvjAfv/HP+DrryE83F+SKt5QRasoSpfm6achJwciIuCll4ItTet57DGIjIRrr4XERJgwAcJa+eSePx/OOw8GDYIzz7RLegCSkvwvr9IUVbSKogSV99+3y1YCgTHwyCOwaBEceyy88grU1rZ8XKiQl2eX3lxwwZ4lPfPnW6VbV9fy8T/9BPvua3vAc+fCFVfAwoUQHx9YuZWGqKJVFCVoGAO33WbnHwPB4sWwfj2cfjpceKEdOv3ss8CcKxA89ZSdk73hhj1pa9faFxNfhsHz8yE1FZ580l7jJ56A0aMDJ6/iGbU6VhQlaCxYACtWNFQk/uSNN+wc5Mkn22HXgQP9tzymI9i5E046CcaO3ZPmclH+3XcwaVLzxx97rN2U4KI92k5Ee8LkZWZmct1113WQpIriGy+/DDExNlrMIYdATY3/6jYG3nwTDjsMUlLscpisLLj6av+dI9D8/e8wp1EIlcGD7QuDh1V9DVi4ELZuDZxsiu9oj7YT4QqTB3YdbeMAAbW1tUREeP5Jp06dytSpUztCTEXxiepq+O9/4cQTITnZWsC+/z6ccop/6i8uhpEj4ayz9qS5/h4lJbaHG6rU1VmXiBkZnq2CDz4YvvjCvkyIeK7jyivt588/B05OxTe0R9vJueiii7jiiiuYNm0at9xyCz///DP7778/U6ZM4Te/+Q2rV68GbCza4447DrBK+pJLLmHGjBkMGzaMJ554IphNULopn35q5xAvuACOOgoGDLDOGPxFjx7w0Ud2btads8+Go4/233kCwbvvwrhx3nutBx1kRwJ27vScv2mTHZY/7bSAiai0gpDr0YrIUcDfgHDgWWPMrEb5g4HngTSgADjPGLPVyasDXN7zNxtjTnDSXwSmA0VO3kXGmMXtlXXGjKZpZ5wBV10F5eVwzDFN8y+6yG47dzb9E3z9ddvk2Lp1Kz/++CPh4eEUFxfz3XffERERweeff84dd9zBW2+91eSYVatW8dVXX1FSUsLo0aO58soriYyMbJsAitIGjj7aKttDDrE9zcsug3vvtcO7w4a1r25jYPt2z873J02C116DdetgxIj2nScQGAN/+Yu9Bgcc4LnMb3/bvKMJ11/+1FP9L5/SekKqRysi4cBTwNFABnC2iGQ0KvYIMNsYMxGYCTzklldhjJnsbCc0Ou5mt7zFAWpCUDj99NMJd8aXioqKOP300xk/fjzXX389y5cv93jMscceS3R0NKmpqfTu3Zvt27d3pMiKQni4dcDgGs699FK7PvRf/2p/3b/8Av3726Hoxpx3nh1unT27/efxJ3V1NpLOYYfZJTw33ODdmURL62jnzLFen4YP97+cSusJtR7tvsA6Y0wWgIi8BpwIrHArkwG4bBS/At7tSAHdaa4HGhfXfH5qatt7sI2Jd1sUd9ddd3HIIYfwzjvvsHHjRmZ46nZjI/+4CA8Pp7YzLS5UOj2vv26Nde67zxpCgR06vusu8Icpgcva2FOPcMAAOPxwq2jvvbf1zh/8xaZN1jtTZCTcf7+V95VX7LPj1lttD785Hn0U3nkHGrs1z8uz87IzZwZOdqV1hJqiTQe2uO1vBaY1KrMEOAU7vHwykCgiKcaYfCBGRDKBWmCWMeZdt+MeEJG7gS+A24wxVQFqQ1ApKioiPT0dgBdffDG4wiiKF558EnbtgocfbpjuFiujzbisjQ8/3AYz98SFF9qe7bffep4CCjRr1tgh87y8hlNMv/7qe7AAEfjhB+v1qn//Pelpadba2ItdpBIEQmro2EduAqaLyCLsvGs24PKRMtgYMxU4B3hcRFwDJ7cDY4B9gF6Ax4iOInK5iGSKSGZeXl4g2xAwbrnlFm6//XamTJmivVQlJFm/3iqI88/3bDGbm2uX/bSVhQthwwbrpMIbJ58Mzz8Pe+/d9vPU11vLZtdWUmKVfEusWgXTp9ulTL/8Yg2fXLQmIs/BB9tPTxEv+/a1o2ZKiGCMCZkN2B/4xG3/duD2ZsonAFu95L0InOYhfQbwv5Zk2XvvvU1jVqxY0SStK9Pd2qt0DPfcY4yIMZs3e86//35jwJi1a9tW/803GxMRYUx+fptF9Movvxjz+ef2e3a2ldN9O+AAY5Yt8358TY0xI0ca06ePMcuXt0+WmhpjEhKMufrqPWk5Ocb83/8Zs3Bh++puD0CmCQF9EkpbqPVoFwAjRWSoiEQBZwFz3QuISKqIuOS+HWuBjIj0FJFoVxngAJy5XRHp53wKcBLQCWN4KErnxxjbWz30UOt0wROXXGKHPdtqFPX739s5Wm/Dxi6qq+Fvf7NLgHyhogLOOQcuvhiqqqxD/kcf3bP96U/WPeKuXd7riIiAF1+09hkZjc08W0lEhI1R696jfftt+OQTu/RHCR1CahTfGFMrItcAn2CX9zxvjFkuIjOxb0lzsT3Sh0TEAN8CLj8vY4FnRKQeOyQ+yxjjMqL6j4ikAQIsBq7oqDYpirKHkhKrHJwl3R7p1w9OOME603c3lvKV9HQ7NNwSERFW0Q4ebNfxenP84OLWW+2w72efWZmio5u6jrz5ZoiNtd//9CcbLefCC21ou3nzrBOJ3/ymde1pjrPOgmXL7DB2WJi1Ns7IaL8SV/xMsLvUobp5Gzqur69vkt4Vqa+v16FjJWh88okxYMx//9u64154wZiXX/a9/F//as9z883GNPfXdsnz+9/7Vm9NjTEHH2yPmTbNmJ49jRk40JiiIt9lay3bthkTFmbM3XcH7hy+gA4dN9lCbeg4pImJiSE/Px97L3VdjDHk5+cTo+NPih+pqvI98Prhh1v3iV6WgXtk3jy480549VXfj/nDH6zv47/8xbvFc2GhHS7OyICHHvJcpjEREfDVV3aYOCvLDjN/801g4r/W1lrL43fftT1b9QYVekhXVxptZerUqSYzM7NBWk1NDVu3bqWysjJIUnUcMTExDBgwQL1FdSGWLrXKIiLCPvwHDrRrODuKt9+2noq+/XZPBJrmqKy0c43G2HW3xx0HCQlNy/38M9xzD3z8sV3a8tZbvtXvor7eelr6z3+sYm/s5KG+3oarO/BA6wSitVRU2DoCFQP28MNtIPfrr7dz02++2fIweCARkYXGrv5QHFTResGTolWUzsoHH9iezi23wN13w4QJ1hjovvvgzDMD57ShrMwuYVm40C6n2bkTtmzx7vHIE4sXWwWXlGR7lldfbXu7YF2dDhhg5b/lFpvXFoVWV2eV7MSJDdNrajr2ZaQt3H67DW5fVGSdXQQbVbRN0aFjRenivPqqjWk6bhxce61VSo88YhXSOefYtaQff+zbGlBfeOMNG3kG4MMP7XrP66+3Svahh1qnZMH6Jp43z/Zo//EPGDXKWi3X11vF8v77dt3sLbe0vdcYHr5Hyb7wgrV43rzZKvRPP21bnR3FQQfZ4eMvvgi2JIpXgj1JHKqbJ2MoRelsPPWUXbM6Y0ZTQ5y6OmP+8x9jhg0zBvxjRLN0qa3rwQftfl6eMe+/b9d3+oOcHGPuvdeYffYxZuVK/9TpTn29Mccfb6/ZiBF2nWpWlv/P408KC83udbyhAGoM1WTToWMv6NCx0tnJybG9v8MOs3Oc3mzbqqth5Urbc2wvt9wCjz1mvTt1Vs9ElZVw/PHw+ed2uPvii4MtUctceaUdXr/88mBLokPHnlBF6wVVtEpXYMkSO2TcEX5v6+rsutG994a5c1suH8pUVNh55QMOCK5hUWdEFW1TdI5WUboY8+dbK9nKSttL9UXJ1tVZ5wd//3vbz/vll7YXff75ba8jVIiNtVbGqmQVf6CKVlG6GLNm2eUurRmsCg+HFSva1xNdudIurzn++LbXoShdEVW0ihLiLF9u5wt9YcMGeO89O1fncgXoK9Onw48/2iUtbeG666ylrvo5UZSGqKJVlBCmrMz6xj3ySDvf2hJ//7tdvnPVVa0/18EH2/MtXNj6Y6uc6M6qZBWlKapoFSXEWLUK/vhHO/QbH28thnv2tBa9zVFaCs89Z+OwDhjQ+vO64pt++23rjz3xRDj77NYfpyjdAVW0ihIibNxoDYnGjYPHH4c1a2z6UUdZxfvpp807T9i2DcaPt2Hi2kKfPnDKKVapt4bcXBvRprHrQkVRLLq8xwu6vEfpSKqrrYLNybFuBG++2RoWuaiqgrFjbQg49/ijocCjj8JNN9me+OjRwZZGCTa6vKcpIRWPVlE6G/n51s3geee1bynIP/8J69bZuo4+uml+dLQdQh482PPxa9ZAcjL07t12GVzU1NjNV7+5L78M++yjSlZRvKFDx4rSRioqrP/dCy7wPfybN4491jr4P+oo72X22ccqUmOsb1t3rr3WBlRv7wBVQQH06gXPPONb+WXLrJFWV1g7qyiBIuQUrYgcJSKrRWSdiNzmIX+wiHwhIktF5GsRGeCWVycii51trlv6UBGZ79T5uohEdVR7lK5JfT1cdBH89JPd/+WX9tU3YoSdh22pV+yyQn7kkT1pK1bYudtLLmm/g4Vevexc7Tff+FZ+4EDrHOOss9p3XkXpyoSUohWRcOAp4GggAzhbRDIaFXsEmG2MmQjMBNxDMVcYYyY72wlu6Q8DjxljRgC7gEsD1gilWzBnjo1SM2uWddRw3nltq2f9emuAtGmTb+Xj4+3c7UMPQV6eTXviCbusxl9+bg8+2M4D19e3XDY52S4lcp9PVhSlISGlaIF9gXXGmCxjTDXwGnBiozIZwJfO96885DdARAQ4FJjjJL0EnOQvgZXuyemnW8cQt9wCY8a0PvSbi9tug08+gahWjLHMmmWX8tx/vx3qnT0bzj3Xf8pu+nRb7/LlzZdbsMCGk6uo8M95FaWrEmqKNh3Y4ra/1UlzZwlwivP9ZCBRRFKc/RgRyRSRn0TkJCctBSg0xrhmtTzVCYCIXO4cn5nn6i4oihvffw9ZWXaI9oQT7Od331mvSK2dH/3xR9szvuUW6NfP9+MyMuCyy2xs1meesT6N27qkxxPTp9vPloaP//53K7v6A1aU5gk1ResLNwHTRWQRMB3IBuqcvMGOWfk5wOMi0qqVfcaYfxljphpjpqbpWJjSiFWrrB/fSxtNPKxcCU8+ad0f+ooxcOONVsHedFPrZbn3XmuJvHkzbNkCEya0vg5vDBkCf/7zHgcWnigrg7fesj179QalKM0Tast7soGBbvsDnLTdGGNycHq0IpIAnGqMKXTysp3PLBH5GpgCvAUki0iE06ttUqeitERenrUMjoqCF15omDdliv1ctAiGDfOtvjfftIZUzz1n511bS79+1gBqr70Co+huvrn5/Icessr2oov8f25F6WqEWo92ATDSsRKOAs4CGsQTEZFUEXHJfTvwvJPeU0SiXWWAA4AVxnrk+Ao4zTnmQuC9gLdE6VJcdpl1JjF3ru3xuTNhgp2jXbzY9/qOOAIefhguvLDtMv3mN4HrTVZWWkWem9s074cfrKK9+GIbr1VRlOYJKUXr9DivAT4BVgJvGGOWi8hMEXFZEc8AVovIGqAP8ICTPhbIFJElWMU6yxizwsm7FbhBRNZh52yf65AGKV2CDz6wCnbmTJg2rWl+TIz12rRoke91unwXt9WIKtBs2QL/93+ew+YlJsIxx8Df/tbxcilKZ0RdMHpBXTAqLoqL7VrRm26CyEjPZS64wHp2+vHH5uuqrbWReB5+2DqgCFWMse4eZ8yAV18NtjRKZ0JdMDYlpHq0ihJqGANJSXD77d6VLMDzz7esZMEuifnqKxtAIJQRsdbH33yzx5p6zhzrFKO8PLiyKUpnQxWtonghKwumTvVt7jXCR7PCzz+3SuzQQ9slWocwfbqdl87Kguxs+N3vrKvJ5l44FEVpiipaRfHC739vnfX7stKrthZOOsk6cGiOzz+3lsIpKc2XCwVcy3u+/toaPlVWwiuvqKJVlNaiilZRPDB3Lvzvf3a9arpH9yYNiYiwzvW//NJ7mdJSmDcPDj/cb2IGlLFjYeFCKCy08WYfewxGjQq2VIrS+VBFq3R6XnzRKsOiIv/UV15uPT2NG2c/fWXy5OYtj7dtg333tcZQnQER617yz3+2jjp++9tgS6QonZNQc1ihKK2itNT6C96+3foMPuOM9tf53HPWyf/XX7dumHTKFOv/uLQUEhKa5o8YYV04dibi4uDnnyE2Vl0tKkpb0R6t0qnZuRNGjrRrWd9/3z91XnUVfPzxHp+/vjJ5srXQXbLEc35ndb4/eLB/AsorSndFFa3SqRkyBL79Fk47zcZlbc+y8Px8WLrUOpH4v/9r/fF77WWDrzcOyg522Dg52RoTKYrSvdChY6XT8u671lNTv37wz39an8FtHd789Vc48USrqFevbptl7YAB3tfSfvEFVFdbAyNFUboX2qNVOiXZ2XD22XDnnXY/IaHtSvbtt2G//ezQ7quvtn/5iqeA6Z99Br167QlAoChK90EVrdIpmTkT6urgrrv2pP31r3DIIb7XUV9vl++ceiqMHw+ZmVbhtoe//c2uka2p2ZNmjF0/e9hhEKb/OEXpdujfXul0rF1rLYN/9zsYOnRPeni4tRRev963eoyxFrUXXWSP69+//bKlpdl1p6tW7Ulbvdr2wI84ov31K4rS+VBFq4QExth1sKtX28gxzXHXXdbK+I9/bJh+/PH2syXr45oaa7AUHm6HjZ9/3n/h5txj07pITrZBBI46yj/nUBSlc6GKVgka9fU2purQoXa9ZnKydZDw6KN78jdtanhMXZ0Nvn7jjdCnT8O8YcOsk4mWFO3zz0NGhl17GxPj3/Who0bZNafuirZvXxsSb+BA/51HUZTOg1odK0EjLMxaDI8YYZVRnz72c+JEm/+//8Epp8A551inFBkZthc6e7b3ZTzHHw+PPGJ7xz16NM2vroYHH7TDxIFYGxoebuV3KdraWuvO8bDDPMujKErXRxWtEhS2brVuE996y3uZqVOtY/+nn7brT4cPh2eesZFvvPVCTzsNSkqsG0VPiu3FF2HzZuv8P1Ceji69FMrK7PcFC6yx1ZtvWtkURel+hFzgdxE5CvgbEA48a4yZ1Sh/MPA8kAYUAOcZY7a65ScBK4B3jTHXOGlfA/0Al2+eI40xO5qTQwO/B46aGttzPf98ePzxlsvv3AlPPGHLpqRYYyhfw9K5U11tvUj162ed+3eES8H77oN77oG8vM4RsUdR2osGfm9KSM3Rikg48BRwNJABnC0iGY2KPQLMNsZMBGYCDzXKvw/41kP15xpjJjtbs0pWCSxffAEFBb7HZE1Ntct5cnLgl19aVrL19TB/flMPTXPm2N7svfcGXsnm51vl+tlnnScsnqIogSGkFC2wL7DOGJNljKkGXgNObFQmA3AFI/vKPV9E9gb6AJ92gKxKG3n9dUhKar2bw4QE6Nmz5XJz59r1sI29NJ15Jnz4YdvcK7aGyko73/zgg50rLJ6iKIEh1BRtOuC+uGOrk+bOEuAU5/vJQKKIpIhIGPAocJOXul8QkcUicpeI5/6MiFwuIpkikpmXl9f2ViheqaqCd96xQdKjowNzjsMOs5bJ7tbHxlhDpaOPDnxvNibGWk8/9ZTtVev6WUXp3oSaovWFm4DpIrIImA5kA3XAVcCH7vO1bpxrjJkAHORs53uq2BjzL2PMVGPM1LS0tMBI38357DNrEXzmmYE7R2IizJhhe7Zg54T32Qdefjlw52zMlCnWecWqVXDggR13XkVRQo9QU7TZgPtqwwFO2m6MMTnGmFOMMVOAO520QmB/4BoR2Yidx71ARGY5+dnOZwnwKnaIWgkChx5qLXADPZx6/PGwZo3dZs+GhQs7dp50yhQ7p9yrV+B67oqidA5CTdEuAEaKyFARiQLOAua6FxCRVGeYGOB2rAUyxphzjTGDjDFDsL3e2caY20QkQkRSnWMjgeOAXzumOUpj4uLsMpeoqMCex+Ul6p134P77bY/26KMDe053Ro+2n5+qtYCidHtCStEaY2qBa4BPgJXAG8aY5SIyU0ROcIrNAFaLyBqs4dMDLVQbDXwiIkuBxdge8r8DIL7SAl99BQ88sGeNaSAZPNj6L05MhI0b7RKbjljO4+KII+COO+DIIzvunIqihCYht442VNB1tP7n7LPt0p6cnLatg20tdXV23WxKig0e0JGKVlG6K7qOtinqGUrpEMrLrRXw+ed3jJIFa2X88svW1aMqWUVRgoUqWqVD+OADO2R8xhkde94DDujY8ymKojQmpOZola7LG29Yt4sHHxxsSRRFUToW7dEqbeK11+Dzz21Ac9cWFwcrVlgDJHdcZgBnn22HcxVFUboTqmiVNvHDD3Y4OD3dxpM94ABreORSsmvW2NisYOdH33zTe2g7RVGUrowqWsVnFi+GP/0J/vpXG03nySc9l/vxR6t4zz0XHn7Y9nR79lSDJEVRuic6R6v4zLPPwkcfQXJy80pz0iS4807bix092oal8yUcnqIoSldEFa3iE+XlNvj6aae1HEEnPt56Y1qxwrparK1VIyhFUbovOnSs+MRbb9lgAJdd5vsxw4fDu+/aZT3x8QETTVEUJaTRHq3iE//+N4wYAdOnt/5YVbKKonRntEertEh9vR0C7t9fDZoURVFaiypapUXCwuDuu4MthaIoSudEh46VZqmutqHmamqCLYmiKErnRBVtN6GtziLefx9OOQU++8y/8iiKonQXVNF2A774wloA/9qGcPfPPmu9P/3f//lfLkVRlO6AKtpuwL33woYN8LvfWcMmX9m8GT75BC65RH0UK4qitJWQU7QicpSIrBaRdSJym4f8wSLyhYgsFZGvRWRAo/wkEdkqIn93S9tbRJY5dT4h0n1sZ+vr4bjj4IQTrGvE557z/dgXXrCfl1wSGNkURVG6AyGlaEUkHHgKOBrIAM4WkYxGxR4BZhtjJgIzgYca5d8HfNso7Z/Ab4GRznaUn0UPWcLC4NZbreOI6dMhM9P3Y7/9Fo44AoYMCZR0iqIoXZ9QW96zL7DOGJMFICKvAScCK9zKZAA3ON+/At51ZYjI3kAf4GNgqpPWD0gyxvzk7M8GTgI+CmA7QoJNm+D7722w9chI+PBD6+DfVz7/HAoKAiefoihKdyCkerRAOrDFbX+rk+bOEuAU5/vJQKKIpIhIGPAocJOHOre2UCcAInK5iGSKSGZeXl4bmxB43n7b+hJuicceg4sugu3b7b5Lya5cCfPnez/OGCgttc4pUlLaLa6iKEq3JtQUrS/cBEwXkUXAdCAbqAOuAj40xmxt7uDmMMb8yxgz1RgzNS0tzT/SBoDkZHj0UZgzx3uZXbusxfDZZ8MAt1lsY+D00+H886GysulxO3bA8cfDoEG2R6soiqK0j1BTtNnAQLf9AU7abowxOcaYU4wxU4A7nbRCYH/gGhHZiJ3HvUBEZjnHD2iuzs5CdTVs2QIHHWSDqv/2t9Yy2BP/+pd15n/jjQ3TRWzIurVr4aFGs9uffAITJ1oF+6c/wWGHBaQZiqIo3YpQU7QLgJEiMlREooCzgLnuBUQk1RkmBrgdeB7AGHOuMWaQMWYIttc72xhzmzEmFygWkf0ca+MLgPc6qD1+5ZFHYOxYyM2FV1+FujobXL22tmG56mobmP2ww2xs2MYcfrg9btYsWLXK9nJvugmOOgpSU2HBArj2WvVrrCiK4g9CStEaY2qBa4BPgJXAG8aY5SIyU0ROcIrNAFaLyBqs4dMDPlR9FfAssA5YTyc0hMrKgvvus8pw0CDrgOIf/7DGTg8+2LDsli12eLlxb9adRx+1c7ZXXGH3q6rg6qutkp0wIWDNUBRF6XaIaatvvi7O1KlTTWZr1sIEEGPgmGOsUl25suGc6z332GDsjZWj62dtrlf63HOwerU1rIqIsEuBFEVR2oOILDTGTA22HKFEqC3vUTzw1lvw8cfWinjAgIZ5f/rTnu81NZCdbS2FExNbrvfSS/0rp6KEKmvWXEV4eBLDh88KtihKN0QVbSdg0SLYay+45hrP+cZY94qFhVBSYt0trlypc6yKAlBfX822bS8SEdFLFa0SFHSwsBPwwAN22DjCy2uRiJ2zffNN2/M9/3xVsorioqQkk/r6Cqqrs6ms3NLyAYriZ1TRhjC//mqNkwBiY5sve/PN1po4MXGPgZOiKFBY+M3u78XFPwVREqW7ooo2RDEGLrsMTjzRWgS3RFgYfPCBHTJWb05KY4wxbN/+X+bPH01W1h+DLU6HUlT0LbGxIwkLi1FFqwQFVbQhys8/WzeJd98N0dG+HRMVZWPHKoo7hYXf88sv+7Fy5TlUV29j8+ZZlJWtDLZYHUJ9fS1FRd/Ts+fhJCTsTXHxvGCLpHRDVNGGKC+8YIeLzzkn2JIoHYExhl9/PZWVKy/0W53l5ev49ddTWbz4IKqqshkz5kWmTVtDeHg869c3dgkeHOrrq1ix4hyWLj02IPWXli6irq6U5OTpJCXtR0nJL9TX+zBEpCh+RBVtCFJRAa+9BqeeCklJwZZGaY6aml3k5Pyb9q5H37nzbXbufJsdO16nrq6sXXUZU8/69beyYEEGBQWfMGTITKZNW0PfvhcSFdWHIUPupqDgQ/LzP27XedpLXV05y5adwI4d/6Wg4EOqq7f7/Ryu+dkePabTo8f+GFNFaekSv5+nI6iu3sm8eYPIz/8g2KIorUQVbQiyYoW1Gr744mBLorTEli2PsGbN5ZSVLW9zHXV1Zaxbdz0REckYU8WuXV+2S6aCgo/ZsuXP9O59JtOmrWXIkLsID98THzE9/VpiY0ewfv0N1NfXNlNT4KitLWHp0qPZtesz0tPtujV3oyV/UVT0DbGxo4iO7ktS0n5A5zWI2r79FaqqtrBz5/vBFkVpJapoQ5C997b+jGfMCLYkSnMYY9ix478AlJevanM9mzY9RFXVFsaNm0N4eAIFBR+2S67s7KeIiurL6NHPER3dr0l+WFgUw4c/Qnn5SnJzn2nXudpCTc0uliw5nKKiHxg79lWGD3+M8PBECgu/9ut5jKmjsPA7kpOnAxAdnU509IBOOU9rjGHbthcAKClpJsalEpKoog0xqqqsxXFMjLpEDHWKi+dTWbkBgPLythkXlZevZcuWv9Cnz3n07HkYPXseQX7+B20eiq6oyKKg4CP69bucsLAor+VSUk4gOflQNmy4m5qaXW06V1uort7B4sWHUFq6mPHj36JPn7MIC4ugR4+DKCz8yq/nKi1dSl1d0W5FC5CUtH+n7NGWli6irGwp0dEDnXa1b3rBG7W1xaxadSkVFRsDUn93RR/lIcYjj8CYMTbEnRLa7NjxX0SiiYzs3aYerTGGdet+T1hYNMOG/RmAlJRjqaraQlnZr22SKSfnn0AY/ftf3mw5EWHEiMeorS1k06aZbTpXa6mqymbx4ulUVKxhwoS5pKaeuDsvOXkG5eWrqKra5rfzuc/PukhK2o/Kyo1+PU9HsG3bC4SFxTB06INAPSUlCwNynk2bHmTbtuepqckLSP3dFXXBGEIYAy++aP0Zx8cHWxqlOYypIy/vDVJSjqGurrxNijY/fy4FBR8xfPhfdw/x9up1NAAFBR+SkNC6MEp1dRXk5j5PWtrJREe3vM4rIWEi/fpdRnb23+nf/wri4ka3ug2eqK0tpbx8OWVlKykv37NVVGQRHh7HxIkfk5x8cINjkpNnAHZOtXfvM/0iR1HRN8TEDCMmZo+DcPd52rS0k/xynkBTV1fJ9u3/ITX1ZHr1+j/AjqY0voaeKCr6gaqqbHr3PqPFshUVWWzd+hh9+lxAUtI+7ZZb2YP2aEOI77+HdevUCCpYGFNHbu7zPhkjFRZ+TXX1Nnr3Ppv4+LGUl6/CmHqfz1VXV8G6dX8gLm7cbmMggOjo/iQkTGmTZemOHa9RW1tA//5X+3zM0KH3ERYW67flPvn5HzFvXn9++WU/Vq++mK1bH6OiIouEhMkMHnwHU6bM86ggEhKm+HWe1ph6Cgu/bTBsbM+zFyKRnWr4OD9/LrW1u+jb92KiotKIiRnms/zr1t3IihVnU1T0Q4tl16+/BZEIhg17sMWySuvQHm0I8cILkJBgl/UoHUtp6VJWr/4tJSU/ExmZyrRpWUREeA+BtH37fwkPTyAl5Thqawuory+nqiqbmJiBPp1v8+aHqazcyKRJXxEWFtkgr1evY9i8eRY1NbuIjOzpU33GGLKznyIublwT5dIcUVG9GTz4LrKybqGg4FN69TrS52Mbk5Pzb9asuZKEhIkMGXIPcXFjiYkZRlhYy4+ZPfO0X7f5/O6UlS2ntragybUID48hIWFKp1K027a9QHT0QHr2PBSApKRpFBZ+2+JxNTWFlJQsAOpZufJ8pk5d4vWeLiz8hp0732LIkJk+jYYorSPkerQicpSIrBaRdSJym4f8wSLyhYgsFZGvRWSAW/ovIrJYRJaLyBVux3zt1LnY2Xp3ZJt8obQU3ngDzjhDh407krq6SrKy7mThwr2prNzAkCH3UlOzk61bH/d6TH19FTt3vkVq6kmEh8cSFzcG8N3yuKIii82bZ9G791n07DmjSX5KyrFAHbt2fepzO0pKfqa0dCHp6VchrYwoMWDAdcTEDGPjxj+1XNgDxhiysv7ImjWX06vXkUye/C2pqScSFzfKJyXrwp/ztJ7mZ11YxxULgra0qTVUVm6loOAT+va9EJFwwMpvAyRsbfZY+9JSz9ChD1FZuYl16673WM6YOtatu57o6IEMHHijn1ugQIgpWrF30lPA0UAGcLaIZDQq9ggw2xgzEZgJPOSk5wL7G2MmA9OA20Skv9tx5xpjJjvbjkC2oy1ER8NLL8Ef/hBsSboPhYXfkJk5kc2bH6R373PZd9+VDBlyD6mpJ7FlyyPU1BR4PK6g4BNqawvp3ftsgFYr2nXrbiAsLJLhwx/xmJ+UtC8RESmtGj7Ozn6K8PBE+vQ53+djXISFRZOaejKlpb+0WvnU11ezatUFbN78AP36Xcb48XOJiEhotQywZ57WH73aoqJviI4eRGzskCZ5SUn7UV9fTlnZsnafpz0YY1r0UrV9+2zA0LfvRbvTkpKmAS0v89m163PCwuIZOPAGBg26lW3bniMv790m5bZtm01p6SKGDXu4wXprxX+ElKIF9gXWGWOyjDHVwGvAiY3KZACuSbSvXPnGmGpjjOuujSb02tYskZF2yHhC6+xflDZgjGHt2utYvHgGxtQxceJnjB37IpGRNhrDkCH3UVdXwubNf/Z4/PbtrxIRkULPnkcAEBnZm4iIZJ8UbW1tCfn5c0lPv8brEJ1IOL16HUVBwUcYU9dindXVeezY8Tp9+lzQ7HB3cyQkTKS+vpKKinU+H1NTU8jSpUexffsrDB16P6NG/atVPdimMkwhPDyp3YrWGONxftZFUtL+gH8dVxhj2Lr17xQV+V7n2rVX8eOP/SguzvRa57ZtL9Cjx3RiY4fvTk9ImIxIFMXFLSnaz0hOnk5YWBRDhtxLQsIU1qz5bQMPXLW1JWzYcAdJSfvRu/dZPsuutI5QU0bpgHvAyK1OmjtLgFOc7ycDiSKSAiAiA0VkqVPHw8aYHLfjXnCGje+S1o6tBZgNG2DmTNi5M9iSdA9KSxeTnf0kffteyj77LKNXr8Mb5CckjKdPn3PJzn6CqqrcBnm1taXk58+ld+/Td8+tighxcWN8WktbWroIMPTocVCz5VJSjqWmZiclJZ4fwu7k5j6HMdWkp1/VYllvxMdPBKCsbKlP5Wtri1i06ECKir5jzJjZDB58Z6uHrBvjr3na8vJV1NTs8KpoY2IGExnZx6+KNjf3Odatu5alS4/yKWBDTs6z5OQ8TX19JUuXHunRLWRR0fdUVKyjX7+G1pFhYdHOPLN3RVtZuZmKijW7XwbDwqIYO/YVamtLWL36st3rtDdvnkV19TZGjHi83b+f4p1QU7S+cBMwXUQWAdOBbKAOwBizxRlSHgFcKCJ9nGPONcZMAA5yNo/jayJyuYhkikhmXl7HrSN74QW4917r41jxTn19DbW1pe2uZ/v2VxCJZPjwP3sdKhsy5F6MqWHTpgcapOfnz6W+vmL3sLELq2hb7tG6FGdi4t7NlrPLOMJaHD42po6cnKdJTj6E+PjGsyy+Ex8/FgintNQ3RZuf/wHl5csZN24Offu2frjaG8nJM6ioWN3kBac1NDc/C/bFKClpP795iCorW866ddfRo8eBhIXFsGzZ8dTU5HstX1y8gLVrr6ZnzyPZZ59lhIfHs2TJEZSVrWhQbtu2FwgPTyAt7bQmdSQlTXMC2nse6t+16wsAevbc8xIZH5/B8OEPk5//P3Jz/01FxUa2bHmUPn3O2z0crQSGUFO02YC72eYAJ203xpgcY8wpxpgpwJ1OWmHjMsCvWKWKMSbb+SwBXsUOUTfBGPMvY8xUY8zUtLQ0vzSoJerr7dzsEUfAQN8MVrsl5eXrWLhwL+bN68+mTbOoq2vbW4kxdezY8V969TqGyMheXsvFxg6nb99Lyc39VwMvOTt2/Jfo6AH06HFgg/JxcWOors6ltrao2fOXlGQSHT2QqKg+zZaLjOxFUtL+5Oc3744xP/8Dqqo2kZ7u+5IeT4SFRRMXN9rnHm1JyULCwmLo1cu/UXf2zNM27/e4trbYa15R0TdERfVrMNzamB499qeiYm2zCtEX6urKWb78TMLDE8nIeJPx49+hqmorv/56KvX11U3KV1fnsXz5qURF9SMj41ViY4czadIXiISzZMnhlJevc9pXyo4db5CWdgbh4U2tI5OSpjnzzJ4dm+za9RmRkX2Ijx/XID09/Vp69jycdeuuZ/XqSxAJY+jQhzzWofiPUFO0C4CRIjJURKKAs4C57gVEJFVEXHLfDjzvpA8QkVjne0/gQGC1iESISKqTHgkch1XCIUFmJmzeDBdcEGxJQpf8/I/45Zd9qKrKoUePA9iw4XZ+/nkM27e/2qq1qwC7dn1FdXUuffqc12LZIUPuAsLYtMla49bUFFBQ8AlpaWey5xa07DGIWt1snSUlmSQmTvVJ1pSUYyktXdhs7876NU4nJaWxKUPrSUiY6LOBUEnJQuLjJ7VrTtazDJNbnKctKPiE77/vycqV5zdRuHZ+9huSk6c3OxS6x3FF+/wGr1v3B8rLlzN27MtER/elR4/9GTPmOYqKvmHt2msauNI0po4VK86munoH48e/tdsmIC5uFJMmfY4xNSxZcigVFRvJy5tDfX0Zfft6XlTvkt+TQZQx9eza9Tk9ex7e5BqIhDF69AuEhUVRWPgVgwbd2sChhxIYQkrRGmNqgWuAT4CVwBvGmOUiMlNETnCKzcAq0DVAH8A1tjcWmC8iS4BvgEeMMcuwhlGfOHO3i7E95H93UJNa5Oef7efBLTt56XYYY9i06UGWLTuW6OjB7L13JhMnfsSkSV8SGZnKypXn8ssv+1FY+L3PdW7f/grh4UmkpBzXYtno6HTS069m27bZlJWtJC/vLYypoU+fs5uUjYsbCzRveVxbW0RFxdpWKVqAgoKPPOYXF2eya9en9O//O78ovPj4iVRWbmyxV25MPaWlv7Q4/N0WWvJ7XFdXzpo1VxIZmcL27a+SmTm5gQFSRcU6qqtzvQ4bu7C/QXi75ml37Hid3Nx/M2jQbQ3WH/fpcy6DBt1Jbu6/2br1b7vTN2z4I4WFXzBq1D+aXLv4+HFMnPgZdXUlLFlyGNnZTxAbO5IePQ7weO6YmKFERqZ6lL+s7FdqavLo1esIL8cOYOzYV0hNPYmBA0MjLnFXJ+QcVhhjPgQ+bJR2t9v3OcAcD8d9Bkz0kF4G+P+J4CeKi2H4cOt2UdlDbW0Jq1ZdxM6db9O799mMHv3s7vnUnj0PYe+9F7B9+ytkZd3B4sUH0bv3WYwZ83KzCqeurpydO99yhuNifJJj0KDbyM39Fxs33k1NTT6xsaNISNirSbmYmKGIRDaraEtKfgFanp91ER8/gejoAeTnf0i/fpc0yNux4w1WrbqYqKj+9O//O5/qa4mEBJdB1K9eH/BglVldXUlAFC3Y37eg4AOqqnKIju7fIG/jxplUVm5g8uSvEYlgxYpzWbToQIYMuZfBg2/fPeTcktOO8PB4EhImtnmetqJiPatX/5akpP0ZMqSpr+ihQ2dSXr6S9etvJC5uFPX1VWzePIt+/S5v8lu6SEyczMSJn7BkyeFUVmYxdOiDXnvle+aZm/Zod+36HIDk5MO8yp+ScuzuFzkl8IRUj7Y7cscdsHatjT+rWMrKVvHLL/uxc+d7DB/+V8aO/U8ToyWRMPr2vYBp09YwaNAd7Njx2u4wYt7Iz3+furpSn4aNXURFpTFgwA3k5c2hsPBrevc+2+PDLywsktjYES0oWmsIlZDgm4ISEXr1OoZduz7dPd9nTB3r19/KihVnkpAwhb33ziQqyj/+V1yWxy0ZRLkc2gdK0Xqbpy0tXcqWLY/Qt+/FJCdPp0ePA5g6dTG9e5/Oxo13sXjxIeTlzSEysvfuofzmcCkqX5ZQuVNfX82KFWchEk5Gxn+bePYCe3+OHTubhIRJrFhxFqtWXUhi4j6MHPlECzLty8SJH5GScjz9+l3abNnExGmUl6+kpqawQfquXZ8RFzdGh4RDCFW0IUCglawxptVGH5s2zWp3APLWUF6+ls2b/8zChfuxYMFYamp2MGnSpwwceH2zc23h4XEMHXo/SUkHsHHj3c2GD9u+/RWiotJb5aIQYODAG4iI6AUYj8PGLlpa4lNSkklMzBCiolJ9PndKyrHU1ZVQVPQ9NTX5LF16NFu2/Jn+/a9k8uQvPcabbSvR0QOIiEhu0SCqpGQhItHExbXdyrk5PM3TGlPP6tWXExnZk+HD/7I7PTIymbFjX2XMmJcoLV3Erl2fkJx8sE9LVZKS9qOursTjy1F9fS2lpUspKVnUZFu37npKSjIZPfo5YmIGe60/PDye8ePnEh4eT1hYNOPGvUVYWHSLcvXocQATJsxt8QVqj+OKBW5yV1FY+G0Da2Ml+ITc0HF34ocf4Pe/t1bH48Y1za+u3kF4eJLPw5yesM4ZriU39xn22ms+iYlNhz0bU1a2ig0bbicp6YDd/lUDQVnZKnbs+C87d76923oyIWFvhg59gL59L/TZ56qIMHz4n1m06AC2bPmrY8TUkOrqnRQUfMyAAdc3MWRqiYiIHowc+QRFRT80G+EmLm4M+fnvU19f47GX0xpDKBfJyYciEsXWrU9QVraUqqpsRo9+tsXeTlsQEeLjJ7bYoy0tXUhCwiSPbfSPHOEkJx/cQNHm5DxNScl8xox5ebcR0Z7yQt++F9CjxwFkZd1G//5X+nQel+OKoqJ5u61zKyu3kJv7HNu2PUdVlXcXh/37X01a2ile813ExAxg771/wZhan/1g+0pS0r6AUFw8f/d8bHHxT9TXl+9eP6uEBqpog8hPP8HCheBpJVFV1TZ+/nkUYWFxDBx4A/37X9lqrz/GGNavv5mcnKcAyM5+kjFjmh9eBcjNfQaA4uIfqarK9WuvyUVl5WYWLpxCfX0VPXocxIgRj5OaelKzPYTm6NHjN6SmnuL09n7XpDeQl/cGxtS2atjYnT59zqVPn3ObLRMXNwZjaqmszGqikGtqdlFZmUW/fr9t1XkjIhJITp5Bfv57REWlM2XKd84DNjDEx09g+/bZGGM89gqNqaek5JcWr0V7sW3+H1VV1udMVtbtJCcf1ux5Y2OHM27cmz6fIzZ2BBERvSgq+p6oqD7k5DzjGJ4ZevY8kqFDHyAiokeT48LC4lr1AhqI/w/YF8C4uDENDKIKCj4Dwls9aqMEFlW0QWTBAhg0CHp7GCHavPlB6urKSUycSlbWrWze/BDp6dcxYMB1Td7ovbFx4z1s3foo/ftfjTG1bN/+EsOHP9Ls8XV1FWzb9hIJCVMoLV3Ezp3vkZ5+hdfybSU393nq66vYZ5/ljrOE9jNs2IP8/PN7bNw4k1Gj/t4gb/v2V4iPn7Db4CcQuPs8bqxo98xrtq5HCzBw4M1ERfVh+PC/tLj+tr0kJEwkJ6eEyspNHv0EV1Ssp66uOGDzsy7c/R7v3PkO9fVVjBr1tF+9F7kMirZvf4nt218iKqovgwbdTr9+lxIbO9Rv5wkkSUnTyM//3+4Xo127Pnd8ZTd9QVCCh87RBpEFC2AfD/GVKys3kZPzDP36XcLkyV+y117zSU6ewaZNM5k3bzDr1t1EVVV20wPd2LTpITZtuo++fS9l5MgnSE+/mvr6SnJzn2/2uLy8N6mt3cXw4Y8SGzuSnTvfbk8TPWJMHdu2PU/Pnkf4TckCxMWNpn//y8nNfYby8rW70ysqsigunhfwXphLuXqa8/PVI5QnevU6nLFjZwdcyULLrhhdLwyeLK/9iZ2n7cGmTQ+QlzeHIUPuIi5uhN/P07//laSlnca4ce+w336bGTbs/k6jZMHOM9fU7KSycsPusHg6bBx6qKINEvn5kJXlWdFu3DgTEAYPtnONSUn7Mn78O+yzz6+kpZ3M1q2PM2/eYJYtO4n8/A+bWE1u2fI4GzbcQe/e5zB69DOIhJGQMIEePQ4iJ+efzVpZ5uQ8Q2zsKJKTZ5CWdiqFhV95jWLTVgoKPqOqagv9+l3m13oBhgy5B5FoNmy4Y3fa9u3/AaB373P8fj53IiJ6EBXVz6uijYkZ7nN82WARHz8e8G55XFq6EJGoJh6H/I1rnra8fAVxcRkMHHhzQM6Tmnoc48a9SVraSQGbcw4kiYnWIKq4+Cdn7XG9GkKFIKpog0RpKZx+elNHFeXlq9m27UXS069sYjwRHz+OsWNfZtq0tQwceBPFxfNYtuxYfvppqLO+cCvZ2U+zfv31pKaeypgxL+2OYQmQnn41lZUbKCj42ItMyygu/pH+/X+HiJCaegrG1JKf/75f256b+28iI1NJTW2/N6PGREX1YdCgm8nLm0NR0U8YY9i+/RWSk2f43RjFE3FxYz0q2tLShW0aNu5oIiISiIkZ3myPNiFhImFhUQGXpWfPIwFh1KhnOuR8nZH4+PGEhcVRXDx/d1g89VsceqiiDRKDB9tA7/vv3zB9w4Z7CAuLZdCg270eGxs7lOHDZ7H//lvIyHiTuLgxbNx4Dz/9NJi1a6+kV69jych4tYnzhtTUk4mK6kt29lMe683JeQaRaPr2vRCw84nR0QPIy/Pf8HF19Xby8+fSp8+FAXt4DhhwI5GRfcjKuoWSkkwqKtbQu3dgh41dxMWNoaxsZQPXe9XVO6ms3BjweU1/kZDg2fLYGENJyS8+rwNuL/37/459911NcvKBLRfupoSFRZCYOHW3onWFxVNCC1W0QaLYg0/0kpLF5OW9zsCB1/vkhCAsLIrevU9j0qRPmTZtHYMG3UK/fpczbtwcj3+2sLAo+vW7nIKCj6moWN8gr66ujO3bX6Z379N3G0u5erU20Hn7o+aADTJtTG1Alqe4iIhIYMiQeygq+o7Vq3+LSJTHCCiBIC5uDHV1RQ1ifpaWtt0QKhjEx0+gomJtk8AN1hCqqMNeGMLCIomLG9kh5+rMuCL5uIfFU0ILVbRBYuxYuOaahmkbNvyRiIhkBgy4sdX1xcYOZ9iwhxg9+plm19327385EEZ29j8bpO/Y8Rp1dcX079/Qwjgt7VSMqaKgoPkoMr5gjCE391l69DjQr0ZQnujX7zJiY0dRVraElJTjiYxMDuj5XLhbHrvYYwgVWAMif2Ets+spL28Ytm3PC0Pn6Jl3F2yAAWt3ofOzoYkq2iCQk2O3kW4v60VFP1JQ8AEDB94aUKUQHZ1OWtrJbNv2fIMeS07O08TFjSMp6TcNyvfocQCRkWl+GT4uKvqOioo1ATGCakxYWCTDhj0MQN++FwX8fC68KdrY2FGdZsmFN1eM1iNU1G6DKSU0cM3JegqLp4QGqmiDQKbt4Oy2ODbGkJV1B5GRfRgw4NqAn79//6uprd3Fjh2vAfYBWlKSSf/+V3gIqxVOaupJFBR8QF1dZbvOm5v7LOHhSR02jJuWdhLTpmWRmtpypB5/ER2dTlhYfCNFu7BT9QJjY4cRFhbXxCDKhsaboHOAIUZ0dDqxsSNJSTnOr+uMFf+hijYILFgA4eEwebLd37Xrc4qKvmHw4Ds9Bnn2N8nJ04mLG0d29lMYY8jJeYawsFivXpPS0k6lrq50d1QQT+Tlvcvy5ac3mJt0p6ZmF3l5b9Knz7kd0kYXHb0mUiSMuLjRuxVtdfV2qqq2dJr5WbAvV/Hx4xv0aI0xAQuNp7Sfvfaa12LAAiV4qKINAgsWWN/GcXH2AbZhwx1ERw9y5k8Dj4iQnn4VpaUL2bXrc7Zvf5Xevc/2OmSdnHwI4eE92LnzLY/5xcXzWbHiLPLy5vDLL/tRVtZ0ecuOHa9SX1/ZIcPGwcYGF7DXoD0eoYKJtTxestt6urIyi9raQlW0IUpkZEqTCFdK6KCKNghcdhnceqv9XlKygJKSTAYP/qNPkT38RZ8+5xMensjKledQX1/WbEzTsLAoUlOPZ+fOudTX1zTIq6zcwq+/nkR0dDoTJnxIXV05ixb9pkEwdttr/jcJCVM6jUFQe4iLG0NV1Sbq6sodQyghIWFKsMVqFfHxE6mtzae6ehvQ+li6iqLsIeQUrYgcJSKrRWSdiNzmIX+wiHwhIktF5GsRGeCW/ouILBaR5SJyhdsxe4vIMqfOJyTIExmnnQbnOE6Kiop+ACAlpePmEQEiIhLp0+cCamp2OgrQg4sqN1JTT6W2toCiom93p9XVlfHrrydQV1fOhAnvk5JyNHvtNY/IyDSWLDmcHTusg/eSkoWUlS1ptUP9zkpcnLWoLi9fTUnJQuLixrQ6IESwiY+fAOxxxWgNoSLVEEpR2kBIKVqxboyeAo4GMoCzRaRx0MtHgNnGmInATOAhJz0X2N8YMxmYBtwmIv2dvH8CvwVGOttRgWxHc2zYAEuXQn293S8u/pHo6MEBi/DRHOnpVyMSSXr6tS0aUfTqdSRhYXHk5dnhY2PqWbnyPEpLl5KR8Trx8fZnio0dxl57/Uhi4lRWrDiDLVseJTf334SFxdK7t/dYrl0Jd8tjGxqv8/UCExKsoi0tXeZ8ugyhOm7URVG6CiGlaIF9gXXGmCxjTDXwGtDYT18G4IpI/pUr3xhTbYypctKjcdomIv2AJGPMT8ZOOM0GTgpoK5rhH/+w1sa1tXa/qGgePXr8pvmDAkR8/Fj23z/Hp+Uv4eFx9Op1NDt3voMx9WzY8Ed27nyXESMeIyWl4XtLZGQKkyZ9Tlra6axffxO5uc+SlnZ6h61lDTaxsSOAMAoLv6S6OqfTzc+C/Q2jotIpK1vqeITqXJbTihJKhJqiTQe2uO1vddLcWQK4Ii6fDCSKSAqAiAwUkaVOHQ8bY3Kc490jOHuqE+f4y0UkU0Qy8/Ly2t0YTyxYYK2No6Ls/GZ1dfbuANTBICoq1eclAWlpp1BdvY21a69j8+aH6Nfvd6Sne16OFB4eQ0bGawwYcAMiYaSnX+VPsUOa8PAYYmKGkpc3B+h8hlAuXK4YKys3Ulu7SxWtorSRUFO0vnATMF1EFgHTgWwctyjGmC3OkPII4EIRaVVcMWPMv4wxU40xU9M8RWNvJ3V1NtC7a/1scfGPAEFVtK3BrtOLIifnKZKTD2XkyCebVdIiYYwY8SgHHLCr2zk6j4sbQ21tIRBGQsLkIEvTNuLjJ1JevmJ3YPGO8nGsKF2NUFO02YB7iJUBTtpujDE5xphTjDFTgDudtMLGZYBfgYOc4wc0V2dHsXq1jdrjUrRFRfMIC4slIWFSMMRpNRERSaSlnUpc3FjGjXvT57BiEREJAZYs9HDN08bHZ3ToumF/kpAwEWNq2LHjVUQid8/bKorSOkJN0S4ARorIUBGJAs4C5roXEJFUEXHJfTvwvJM+QERine89gQOB1caYXKBYRPZzrI0vAN7rmOY0pLFHqOLieSQm7tOp4mCOGfMS++yzjMjIXsEWJaRxKdrO3At0WR7n53/khGNTQyhFaQshpWiNMbXANcAnwErgDWPMchGZKSInOMVmAKtFZA3QB3jASR8LzBeRJcA3wCPGmGVO3lXAs8A6YD3wUUe0pzEnnAAffQSjR0NdXQWlpb90mmFjF2FhkQ1i3CqecSnazjo/CxAXNxqRSKBO52cVpR1EtFykYzHGfAh82Cjtbrfvc4A5Ho77DJjopc5MIOgLAJOT4SjHQLewcCHG1AbN4lgJLElJ+zFkyEz69Dkn2KK0mbCwKOLixlJWtrRT98wVJdiEVI+2K1NTAw8+CKsc74TFxfMAV4grpasRFhbBkCF3dfohdhsyTz1CKUp7UEXbQSxbBnfeCUuW2P3i4h+JiRnuU4B3RQkWycmHOuHX1BBKUdqKKtoOYqH1Lc/Uqdb3bzAdVSiKr/TtexG/+U0O4eExwRZFUTotITdH21VZuhQSE2HoUKis3EhNzfZOZwildD+sob7GOFWU9qA92g5i6VIYPx7CwjqfowpFURSl7aii7SDWrIGJjk10UdE8wsMTNBKKoihKN0CHjjuIzZuhrMx+t44q9iUsTC+/oihKV0d7tB1EZKRdR1tXV0Zp6RIdNlYURekmqKLtAF57Da67zgYVKC5eANSpxbGiKEo3QRVtB/D++/DeexAero4qFEVRuhuqaDuApUv3GEIVF/9IbOzoTu8xSFEURfENVbQBprraul2cMME6qigu/kmHjRVFUboRqmgDzKpVUFtre7QVFeuoqdmphlCKoijdCFW0AWbnThgwwPZo1VGFoihK90MXcgaYQw+FLVvs99Wr5xEenkR8fEZwhVIURVE6DO3RdiDFxfNIStoPEb3siqIo3YWQe+KLyFEislpE1onIbR7yB4vIFyKyVES+FpEBTvpkEZknIsudvDPdjnlRRDaIyGJnm9xR7TnwQHjySaitLaasbJkOGyuKonQzQkrRikg48BRwNJABnC0ijcdZHwFmG2MmAjOBh5z0cuACY8w44CjgcRFJdjvuZmPMZGdbHMBm7GbnTvjhB2t5XFz8M2DU4lhRFKWbEVKKFtgXWGeMyTLGVAOvASc2KpMBfOl8/8qVb4xZY4xZ63zPAXYAaR0itReWLbOfEydCcfFPgJCYuG8wRVIURVE6mFBTtOnAFrf9rU6aO0uAU5zvJwOJIpLiXkBE9gWigPVuyQ84Q8qPiUi0p5OLyOUikikimXl5ee1pB2AdVYDL4vgn4uLGEBmZ3O56FUVRlM5DqClaX7gJmC4ii4DpQDZQ58oUkX7Ay8DFxph6J/l2YAywD9ALuNVTxcaYfxljphpjpqaltb8zvGwZpKVB796GkpL5JCVNa3ediqIoSuci1Jb3ZAMD3fYHOGm7cYaFTwEQkQTgVGNMobOfBHwA3GmM+cntmFzna5WIvIBV1gFn+HA480yoqtrgOKpQ/8aKoijdjVBTtAuAkSIyFKtgzwLOcS8gIqlAgdNbvR143kmPAt7BGkrNaXRMP2NMrogIcBLwa6AbAnD77fZz+3ar8xMTtUerKIrS3QipoWNjTC1wDfAJsBJ4wxizXERmisgJTrEZwGoRWQP0AR5w0s8ADgYu8rCM5z8isgxYBqQC9we6LbW1UO8MXBcXzycsLI74+PGBPq2iKIoSYogxJtgyhCRTp041mZmZbT7+rbfgwgthwQIoL9+PsLBopkz5xo8SKoqihB4istAYMzXYcoQSIdWj7UosXQoVFTBoUBWlpYvUEEpRFKWbooo2QCxdCqNGQV3dYoypVkWrKIrSTVFFGyCWLnWtn50PoBbHiqIo3RRVtAGgtBSysvZ4hIqKSic6urHfDUVRFKU7oIo2ANTUwN13w5FH2h6tDhsriqJ0X1TRBoCePeFPf4LJk/OorMzSYWNFUZRuTKg5rOgSZGVBr15QW+uan9UeraIoSndFe7QB4OKL4ZhjoKRkPhBOYuLewRZJURRFCRKqaP2MMdbieNIkawiVkDCB8PD4YIulKIqiBAlVtH5m61YoLITx4+spLv5Z/RsriqJ0c1TR+hlXDNpRozZTV1es87OKoijdHFW0fmbZMvs5aND3gDqqUBRF6e6o1bGfOflkSE8H+J7w8B7ExY0OtkiKoihKEFFF62dGj7ZbZuZ8kpL2QUQHDRRFUbozqgUCQF1dGaWlS3XYWFEURVFFGwhKShYC9WpxrCiKooSeohWRo0RktYisE5HbPOQPFpEvRGSpiHwtIgOc9MkiMk9Eljt5Z7odM1RE5jt1vi4iUYFsw56IPapoFUVRujshpWhFJBx4CjgayADOFpGMRsUeAWYbYyYCM4GHnPRy4AJjzDjgKOBxEUl28h4GHjPGjAB2AZcGsh3FxT8REzOUqKi0QJ5GURRF6QSElKIF9gXWGWOyjDHVwGvAiY3KZABfOt+/cuUbY9YYY9Y633OAHUCaiAhwKDDHOeYl4KRANsJG7NH5WUVRFCX0FG06sMVtf6uT5s4S4BTn+8lAooikuBcQkX2BKGA9kAIUGmNqm6nTddzlIpIpIpl5eXltakBl5Vaqq7N12FhRFEUBQk/R+sJNwHQRWQRMB7KBOlemiPQDXgYuNsbUt6ZiY8y/jDFTjTFT09LaNuxrAwmghlCKoigKEHrraLOBgW77A5y03TjDwqcAiEgCcKoxptDZTwI+AO40xvzkHJIPJItIhNOrbVKnPykuno9IFImJUwJ1CkVRFKUTEWo92gXASMdKOAo4C5jrXkBEUmWPF4jbgeed9CjgHayhlGs+FmOMwc7lnuYkXQi8F6gGxMQMo2/fiwgLiw7UKRRFUZROREgpWqfHeQ3wCbASeMMYs1xEZorICU6xGcBqEVkD9AEecNLPAA4GLhKRxc422cm7FbhBRNZh52yfC1Qb0tOvYPToZwJVvaIoitLJENvhUxozdepUk5mZGWwxFEVROhUistAYMzXYcoQSIdWjVRRFUZSuhipaRVEURQkgqmgVRVEUJYCoolUURVGUAKKKVlEURVECiCpaRVEURQkgqmgVRVEUJYDoOloviEgesKmNh6cCO/0oTmdB29296K7thu7bdl/aPdgYozFC3VBFGwBEJLM7LtjWdncvumu7ofu2vbu2u73o0LGiKIqiBBBVtIqiKIoSQFTRBoZ/BVuAIKHt7l5013ZD9217d213u9A5WkVRFEUJINqjVRRFUZQAoopWURRFUQKIKlo/IyJHichqEVknIrcFW55AISLPi8gOEfnVLa2XiHwmImudz57BlDEQiMhAEflKRFaIyHIR+b2T3qXbLiIxIvKziCxx2v0nJ32oiMx37vfXRSQq2LIGAhEJF5FFIvI/Z7/Lt1tENorIMhFZLCKZTlqXvs8DhSpaPyIi4cBTwNFABnC2iGQEV6qA8SJwVKO024AvjDEjgS+c/a5GLXCjMSYD2A+42vmNu3rbq4BDjTGTgMnAUSKyH/Aw8JgxZgSwC7g0eCIGlN8DK932u0u7DzHGTHZbO9vV7/OAoIrWv+wLrDPGZBljqoHXgBODLFNAMMZ8CxQ0Sj4ReMn5/hJwUkfK1BEYY3KNMb8430uwD990unjbjaXU2Y10NgMcCsxx0rtcuwFEZABwLPCssy90g3Z7oUvf54FCFa1/SQe2uO1vddK6C32MMbnO921An2AKE2hEZAgwBZhPN2i7M3y6GNgBfAasBwqNMbVOka56vz8O3ALUO/spdI92G+BTEVkoIpc7aV3+Pg8EEcEWQOmaGGOMiHTZtWMikgC8BfzBGFNsOzmWrtp2Y0wdMFlEkoF3gDHBlSjwiMhxwA5jzEIRmRFkcTqaA40x2SLSG/hMRFa5Z3bV+zwQaI/Wv2QDA932Bzhp3YXtItIPwPncEWR5AoKIRGKV7H+MMW87yd2i7QDGmELgK2B/IFlEXC/sXfF+PwA4QUQ2YqeCDgX+RtdvN8aYbOdzB/bFal+60X3uT1TR+pcFwEjHIjEKOAuYG2SZOpK5wIXO9wuB94IoS0Bw5ueeA1YaY/7qltWl2y4iaU5PFhGJBY7Azk9/BZzmFOty7TbG3G6MGWCMGYL9P39pjDmXLt5uEYkXkUTXd+BI4Fe6+H0eKNQzlJ8RkWOwczrhwPPGmAeCK1FgEJH/AjOwYbO2A/cA7wJvAIOwIQbPMMY0Npjq1IjIgcB3wDL2zNndgZ2n7bJtF5GJWOOXcOwL+hvGmJkiMgzb0+sFLALOM8ZUBU/SwOEMHd9kjDmuq7fbad87zm4E8Kox5gERSaEL3+eBQhWtoiiKogQQHTpWFEVRlACiilZRFEVRAogqWkVRFEUJIKpoFUVRFCWAqKJVFEVRlACiilZRFEVRAogqWkVRFEUJIP8PjShzdBJkp3kAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.9 for train and test data\n",
    "# n =20 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0.9, n = 20 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, very close to matlab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 178,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b2b829220>"
      ]
     },
     "execution_count": 178,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.9 for train and test data\n",
    "# n =50 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0.9, n = 50 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, very close to matlab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 190,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b2eecbac0>"
      ]
     },
     "execution_count": 190,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.9 for train and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0.9, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, very close to matlab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 213,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b28f5fdc0>"
      ]
     },
     "execution_count": 213,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0 for train and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, very close to matlab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b2c5a00d0>"
      ]
     },
     "execution_count": 225,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.25 for train and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0.25, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b330cb0d0>"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.5 for train and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('Training and test Accuracy for eta = 0.1, alpha = 0.5, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 304,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b2a9a2070>"
      ]
     },
     "execution_count": 304,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.9 for 1/2 train data and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('1/2 Training and test Accuracy for eta = 0.1, alpha = 0.9, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 322,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f8b3148c550>"
      ]
     },
     "execution_count": 322,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#######################################\n",
    "#\n",
    "# Plotting eta =0.1, alpha =0.9 for 1/4 train data and test data\n",
    "# n =100 hidden layers, 50 epochs\n",
    "#\n",
    "########################################\n",
    "\n",
    "x = np.arange(0, epo) \n",
    "y = accuracy_test \n",
    "z = accuracy_training\n",
    "plt.plot(x,y,'y-')\n",
    "plt.plot(x,z,'b--')\n",
    "plt.title('1/4 Training and test Accuracy for eta = 0.1, alpha = 0.9, n = 100 for each Epoch')\n",
    "plt.legend(['Test', 'Train']) #just need extra brackets, "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}