congressional voting predictions.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "# One of the way to build keras model is by using sequential API. \n",
    "# Sequential API allows models to be built layer by layer\n",
    "from keras.models import Sequential\n",
    "# Since layers are connected, thus we use Dense class\n",
    "from keras.layers import Dense\n",
    "from keras import optimizers\n",
    "from keras import metrics\n",
    "from sklearn.metrics import classification_report, confusion_matrix\n",
    "import matplotlib.pyplot as plt\n",
    "from matplotlib import pyplot\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "from sklearn.preprocessing import LabelEncoder\n",
    "from keras.optimizers import Adam"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"votingrecords.csv\", header = None)\n",
    "df1 = df.apply(LabelEncoder().fit_transform)\n",
    "# The sklearn library, fit and transform the data, and then \n",
    "# replace the existing text/categorical data with the new encoded data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "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",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   0   1   2   3   4   5   6   7   8   9   10  11  12  13  14  15  16\n",
       "0   0   0   1   1   0   1   1   0   0   0   0   0   0   1   1   1   1\n",
       "1   1   0   1   0   1   1   1   0   0   0   0   0   1   1   1   0   1\n",
       "2   0   1   1   1   0   0   0   1   1   1   0   1   0   0   0   1   1\n",
       "3   0   1   1   1   0   0   0   1   1   1   0   0   0   0   0   1   1\n",
       "4   0   1   0   1   0   0   0   1   1   1   1   0   0   0   0   1   1"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df1.head() # tranformed data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 0 is for Democrats and 1 is for republican"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "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",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "      <th>9</th>\n",
       "      <th>10</th>\n",
       "      <th>11</th>\n",
       "      <th>12</th>\n",
       "      <th>13</th>\n",
       "      <th>14</th>\n",
       "      <th>15</th>\n",
       "      <th>16</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>democrat</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>republican</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>democrat</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>democrat</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>democrat</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>n</td>\n",
       "      <td>y</td>\n",
       "      <td>y</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           0  1  2  3  4  5  6  7  8  9  10 11 12 13 14 15 16\n",
       "0    democrat  n  y  y  n  y  y  n  n  n  n  n  n  y  y  y  y\n",
       "1  republican  n  y  n  y  y  y  n  n  n  n  n  y  y  y  n  y\n",
       "2    democrat  y  y  y  n  n  n  y  y  y  n  y  n  n  n  y  y\n",
       "3    democrat  y  y  y  n  n  n  y  y  y  n  n  n  n  n  y  y\n",
       "4    democrat  y  n  y  n  n  n  y  y  y  y  n  n  n  n  y  y"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import defaultdict\n",
    "df = pd.read_csv(\"votingrecords.csv\", header = None)\n",
    "d = defaultdict(LabelEncoder)\n",
    "# converting all the columns to dictionary of \"df\"\n",
    "# and then encoding them by using fit.transform()\n",
    "fit = df.apply(lambda x: d[x.name].fit_transform(x))\n",
    "# Inverse the encoded\n",
    "df2 = fit.apply(lambda x: d[x.name].inverse_transform(x))\n",
    "df2.head()\n",
    "# To verify if the tranformation happened as intended by using \n",
    "# inverse_transform()."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# validating\n",
    "df.equals(df2)\n",
    "# The original data frame and the \"df2\" which was inverse \n",
    "# dataframe are same."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 73,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# validating\n",
    "df1.equals(fit)\n",
    "# the transformed dataframe's are also same"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# validating\n",
    "df.shape == df1.shape\n",
    "# both original and transformed data have same dimensions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The pandas is now changed to numpy array <class 'numpy.ndarray'>\n"
     ]
    }
   ],
   "source": [
    "df1 = df1.values\n",
    "print('The pandas is now changed to numpy array',type(df1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(232, 17)\n"
     ]
    }
   ],
   "source": [
    "# \"df1\" has 232 observation and 17 variables\n",
    "# Features(2nd to last variables) and Target(1st variables) are assigned \n",
    "# to X and Y.\n",
    "print(df1.shape)\n",
    "Y = df1[:,0]\n",
    "X = df1[:,1:17]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 2 hidden layers with 12 and 8 hidden nuerons - 150 epochs - 0.0001 learning rate ==========================="
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "232/232 [==============================] - 0s 1ms/step\n",
      "\n",
      "acc: 84.48%\n",
      "\n",
      "loss: 45.58%\n",
      "\n",
      "================\n",
      "Confusion Matrix\n",
      "================\n",
      "True negatives:  124\n",
      "False negatives:  36\n",
      "False positives:  0\n",
      "True positives:  72\n"
     ]
    }
   ],
   "source": [
    "# np.random.seed(7) will produce the same set of network weights \n",
    "np.random.seed(7)\n",
    "\n",
    "# building keras model using sequential API\n",
    "model = Sequential()\n",
    "\n",
    "# adding 1st hidden layer which has 12 nuerons, input layer has 8 nuerons,\n",
    "# for 8 variables in 'X' dataset. The inputs are specifically mentioned \n",
    "# by using \"input_dim\" argument.\n",
    "# activation function \"relu\" is used.\n",
    "# The addition of product of inputs and the random weights of all the layers\n",
    "# below will passed through the respective activation function.\n",
    "model.add(Dense(12, input_dim=16, activation='relu'))\n",
    "\n",
    "# adding 2nd hidden layer with 8 nuerons and passing it thtough relu function.\n",
    "# The \"relu\" will predict the output from the 2nd hidden layer in 1 or 0. \n",
    "model.add(Dense(8, activation='relu'))\n",
    "\n",
    "# adding 3rd hidden/output layer which has 1 neurons, the activation\n",
    "# function used here is \"sigmoid\" which predicts the targets, ie. \"Y\"\n",
    "# by values between 1 and 0\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "opt = Adam(lr=0.0001)\n",
    "# Adam optimizer default learning rate is 0.001.\n",
    "\n",
    "# after adding all the layers to the model, compiling them together with\n",
    "# additional properties such as loss \"binary_crossentropy\" , \n",
    "# gradient descent optimizer called \"adam\". The predicted observations \n",
    "# from the sigmoid activation function will be used to compute \n",
    "# loss = -{(y\\log(p) + (1 - y)\\log(1 - p))}, where log is the natural log, \n",
    "# y is binary indicator (0 or 1) and p is predicted probability. Further \n",
    "# the weights will be reduced by optimizer till the loss is low or until\n",
    "# the specified epochs are completed. Metric allows to monitor the model while \n",
    "# training. accuracy value and loss is generated for each epoch in verbose output.\n",
    "# compiling the model involves Tensorflow at backend using which the weights are\n",
    "# evaluated and the predictions are made.\n",
    "model.compile(loss='binary_crossentropy', optimizer=opt, metrics=['accuracy'])\n",
    "\n",
    "# The model created above will be now trained with data using fit() function.\n",
    "# For every epoch the loss will be decreased to its minimum.\n",
    "# verbose = 0 turns off the output progress bars during training\n",
    "one_fifty_history = model.fit(X, Y, epochs=150, verbose=0)\n",
    "\n",
    "# evaluate() function is used to check how the model will perform\n",
    "# on the test data set, here we will evaluate with the same dataset passed in\n",
    "# while training the model.When the score are printed in the output\n",
    "# it shows the values of overall [loss, accuracy] of the model\n",
    "scores = model.evaluate(X, Y)\n",
    "\n",
    "# predict() function is used for predicting the targets of the new dataset if any.\n",
    "# here we will just predict for the features dataset we have and check its accuracy\n",
    "# in the output below\n",
    "Y_predict = model.predict(X)\n",
    "\n",
    "# model.metrics_names[0:2] has first element \"loss\" and 2nd \"accuracy\"\n",
    "# where as scores 2nd element is loss and we are printing the same.\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[1], scores[1]*100))\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[0], scores[0]*100))\n",
    "\n",
    "# round() function rounds off the number to the nearest integer. \n",
    "rounded = [round(i[0]) for i in Y_predict]\n",
    "y_pred = np.array(rounded,dtype='int64')\n",
    "\n",
    "#confusion matrix is used to evaluate the quality of the output of the model.\n",
    "#The diagonal elements represent the number of points for which the predicted \n",
    "#label is equal to the true label, while off-diagonal elements are those that \n",
    "#are mislabeled by the model.\n",
    "print('')\n",
    "print('================')\n",
    "print('Confusion Matrix')\n",
    "print('================')\n",
    "CM = confusion_matrix(Y, y_pred)\n",
    "print('True negatives: ',CM[0,0])\n",
    "print('False negatives: ',CM[1,0])\n",
    "print('False positives: ',CM[0,1])\n",
    "print('True positives: ',CM[1,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "In above NN, the model predicted 36 actual Republicans as Democrats."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.45578636280421553, 0.8448275862068966]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(scores)\n",
    "acc_onefifty = one_fifty_history .history['acc']\n",
    "loss_onefifty= one_fifty_history .history['loss']\n",
    "epoch = range(0,151)\n",
    "ax = plt.subplot(111)\n",
    "ax.plot(acc_onefifty, label=\"Accuracy\")\n",
    "plt.legend()\n",
    "ax.plot(loss_onefifty, label=\"Loss\")\n",
    "plt.legend()\n",
    "plt.xlabel(\"Number of epochs\")\n",
    "plt.ylabel(\"loss/accuracy values\")\n",
    "plt.title(\"Change in Loss and Accuracy with respect to number of epochs\",fontweight='bold')\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.xticks(np.arange(min(epoch), max(epoch)+1,15))\n",
    "plt.show()\n",
    "# The plot shows the model has increasein accuracy from approximately \n",
    "# 0.5 to 0.84(i.e 50% to 84% in 150epoch)\n",
    "# and loss reduces from  0.8 to approximately 0.45 during \n",
    "# the training process for 150 epochs. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = plt.subplot(1,1,1)\n",
    "plt.hist(y_pred,alpha=0.5,histtype='stepfilled',label=\"prediction\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.hist(Y,alpha=0.5,histtype='stepfilled', label=\"Y_observed\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.xlabel(\"values of observations\")\n",
    "plt.ylabel(\"number of obseravtions\")\n",
    "plt.title(\"Comparisons of observed and predicted \\n values for 150 epochs\")\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.show()\n",
    "plt.rcParams['figure.figsize'] = [10,5]\n",
    "# The prediction which are closer to 1's and 0's are plotted over \n",
    "# Y_observed. 84.4% are predicted close/equal to actual values."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 3 hidden layers with 8,16 and then 32 hidden nuerons - 10 epochs - 0.1 learning rate ==========================="
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "232/232 [==============================] - 0s 1ms/step\n",
      "\n",
      "acc: 99.57%\n",
      "\n",
      "loss: 2.11%\n",
      "\n",
      "================\n",
      "Confusion Matrix\n",
      "================\n",
      "True negatives:  123\n",
      "False negatives:  0\n",
      "False positives:  1\n",
      "True positives:  108\n"
     ]
    }
   ],
   "source": [
    "np.random.seed(7)\n",
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(8, input_dim=16, activation='relu'))\n",
    "\n",
    "model.add(Dense(16, activation='relu'))\n",
    "\n",
    "model.add(Dense(32, activation='relu'))\n",
    "\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "opt = Adam(lr=0.1)\n",
    "model.compile(loss='binary_crossentropy', optimizer= opt, metrics=['accuracy'])\n",
    "\n",
    "ten_history = model.fit(X, Y, epochs=10, verbose=0)\n",
    "\n",
    "scores = model.evaluate(X, Y)\n",
    "\n",
    "Y_predict = model.predict(X)\n",
    "\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[1], scores[1]*100))\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[0], scores[0]*100))\n",
    "\n",
    "rounded = [round(i[0]) for i in Y_predict]\n",
    "y_pred = np.array(rounded,dtype='int64')\n",
    "\n",
    "print('')\n",
    "print('================')\n",
    "print('Confusion Matrix')\n",
    "print('================')\n",
    "CM = confusion_matrix(Y, y_pred)\n",
    "print('True negatives: ',CM[0,0])\n",
    "print('False negatives: ',CM[1,0])\n",
    "print('False positives: ',CM[0,1])\n",
    "print('True positives: ',CM[1,1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1 actual democrat is predicted as republican"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.021094730155992097, 0.9956896551724138]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(scores)\n",
    "acc_ten = ten_history .history['acc']\n",
    "loss_ten= ten_history .history['loss']\n",
    "epoch = range(0,11)\n",
    "ax = plt.subplot(111)\n",
    "ax.plot(acc_ten, label=\"Accuracy\")\n",
    "plt.legend()\n",
    "ax.plot(loss_ten, label=\"Loss\")\n",
    "plt.legend()\n",
    "plt.xlabel(\"Number of epochs\")\n",
    "plt.ylabel(\"loss/accuracy values\")\n",
    "plt.title(\"Change in Loss and Accuracy with respect to number of epochs\",fontweight='bold')\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.xticks(np.arange(min(epoch), max(epoch)+1,1))\n",
    "plt.show()\n",
    "# The plot shows the model has increasein accuracy from approximately \n",
    "# 0.8 to 0.99(i.e 80% to 99% in 10epoch)\n",
    "# and loss reduces from  0.75 to approximately 0.02 during \n",
    "# the training process for 10 epochs. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = plt.subplot(1,1,1)\n",
    "plt.hist(y_pred,alpha=0.5,histtype='stepfilled',label=\"prediction\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.hist(Y,alpha=0.5,histtype='stepfilled', label=\"Y_observed\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.xlabel(\"values of observations\")\n",
    "plt.ylabel(\"number of obseravtions\")\n",
    "plt.title(\"Comparisons of observed and predicted \\n values for 10 epochs\")\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.show()\n",
    "plt.rcParams['figure.figsize'] = [10,5]\n",
    "# The prediction which are closer to 1's and 0's are plotted over \n",
    "# Y_observed. 99% are predicted close/equal to actual values"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### 1 hidden layers with 128 hidden nuerons - 500 epochs - 0.01 learning rate ==========================="
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "232/232 [==============================] - 0s 2ms/step\n",
      "\n",
      "acc: 100.00%\n",
      "\n",
      "loss: 0.00%\n",
      "\n",
      "================\n",
      "Confusion Matrix\n",
      "================\n",
      "True negatives:  124\n",
      "False negatives:  0\n",
      "False positives:  0\n",
      "True positives:  108\n"
     ]
    }
   ],
   "source": [
    "np.random.seed(7)\n",
    "model = Sequential()\n",
    "\n",
    "model.add(Dense(128, input_dim=16, activation='relu'))\n",
    "\n",
    "model.add(Dense(1, activation='sigmoid'))\n",
    "\n",
    "opt = Adam(lr=0.1)\n",
    "model.compile(loss='binary_crossentropy', optimizer= opt, metrics=['accuracy'])\n",
    "\n",
    "five_history = model.fit(X, Y, epochs=500, verbose=0)\n",
    "\n",
    "scores = model.evaluate(X, Y)\n",
    "\n",
    "Y_predict = model.predict(X)\n",
    "\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[1], scores[1]*100))\n",
    "print(\"\\n%s: %.2f%%\" % (model.metrics_names[0], scores[0]*100))\n",
    "\n",
    "rounded = [round(i[0]) for i in Y_predict]\n",
    "y_pred = np.array(rounded,dtype='int64')\n",
    "\n",
    "print('')\n",
    "print('================')\n",
    "print('Confusion Matrix')\n",
    "print('================')\n",
    "CM = confusion_matrix(Y, y_pred)\n",
    "print('True negatives: ',CM[0,0])\n",
    "print('False negatives: ',CM[1,0])\n",
    "print('False positives: ',CM[0,1])\n",
    "print('True positives: ',CM[1,1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[0.0014687232198474672, 1.0]\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(scores)\n",
    "acc_five = five_history .history['acc']\n",
    "loss_five = five_history .history['loss']\n",
    "epoch = range(0,11)\n",
    "ax = plt.subplot(111)\n",
    "ax.plot(acc_five, label=\"Accuracy\")\n",
    "plt.legend()\n",
    "ax.plot(loss_five, label=\"Loss\")\n",
    "plt.legend()\n",
    "plt.xlabel(\"Number of epochs\")\n",
    "plt.ylabel(\"loss/accuracy values\")\n",
    "plt.title(\"Change in Loss and Accuracy with respect to 500 number of epochs\",fontweight='bold')\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.xticks(np.arange(min(epoch), max(epoch)+1,100))\n",
    "plt.show()\n",
    "# The plot shows the model has increase in accuracy from approximately \n",
    "# 0.54 to 1 (i.e 54% to 100% in 10epoch)\n",
    "# and loss reduces from  0.67 to approximately 0 during \n",
    "# the training process for 10 epochs. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ax = plt.subplot(1,1,1)\n",
    "plt.hist(y_pred,alpha=0.5,histtype='stepfilled',label=\"prediction\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.hist(Y,alpha=0.5,histtype='stepfilled', label=\"Y_observed\")\n",
    "ax.legend()\n",
    "ax.legend(loc='best', bbox_to_anchor=(0.78,0.92))\n",
    "plt.xlabel(\"values of observations\")\n",
    "plt.ylabel(\"number of obseravtions\")\n",
    "plt.title(\"Comparisons of observed and predicted \\n values for 10 epochs\")\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['top'].set_visible(False)\n",
    "plt.show()\n",
    "plt.rcParams['figure.figsize'] = [10,5]\n",
    "# all y_pred matches with Y_observed. 100% accuracy."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "references:\n",
    "\n",
    "https://stackoverflow.com/questions/24458645/label-encoding-across-multiple-columns-in-scikit-learn encoding categorical to binary\n",
    "\n",
    "https://www.w3schools.com/python/python_lambda.asp lamda\n",
    "\n",
    "https://medium.com/@contactsunny/label-encoder-vs-one-hot-encoder-in-machine-learning-3fc273365621 encoding vs one hot encoding\n",
    "\n",
    "https://machinelearningmastery.com/understand-the-dynamics-of-learning-rate-on-deep-learning-neural-networks/ for different learning rate"
   ]
  }
 ],
 "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": "3.4.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}