LinearRegressionData.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 132,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatetimeIndex(['2018-02-18 21:00:00', '2018-02-18 22:00:00',\n",
       "               '2018-02-18 23:00:00', '2018-02-19 00:00:00',\n",
       "               '2018-02-19 01:00:00', '2018-02-19 02:00:00',\n",
       "               '2018-02-19 03:00:00', '2018-02-19 04:00:00',\n",
       "               '2018-02-19 05:00:00', '2018-02-19 06:00:00',\n",
       "               '2018-02-19 07:00:00', '2018-02-19 08:00:00',\n",
       "               '2018-02-19 09:00:00', '2018-02-19 10:00:00',\n",
       "               '2018-02-19 11:00:00', '2018-02-19 12:00:00',\n",
       "               '2018-02-19 13:00:00', '2018-02-19 14:00:00',\n",
       "               '2018-02-19 15:00:00', '2018-02-19 16:00:00',\n",
       "               '2018-02-19 17:00:00', '2018-02-19 18:00:00',\n",
       "               '2018-02-19 19:00:00', '2018-02-19 20:00:00',\n",
       "               '2018-02-19 21:00:00', '2018-02-19 22:00:00',\n",
       "               '2018-02-19 23:00:00', '2018-02-20 00:00:00',\n",
       "               '2018-02-20 01:00:00', '2018-02-20 02:00:00',\n",
       "               '2018-02-20 03:00:00', '2018-02-20 04:00:00',\n",
       "               '2018-02-20 05:00:00', '2018-02-20 06:00:00',\n",
       "               '2018-02-20 07:00:00', '2018-02-20 08:00:00',\n",
       "               '2018-02-20 09:00:00', '2018-02-20 10:00:00',\n",
       "               '2018-02-20 11:00:00', '2018-02-20 12:00:00',\n",
       "               '2018-02-20 13:00:00', '2018-02-20 14:00:00',\n",
       "               '2018-02-20 15:00:00', '2018-02-20 16:00:00',\n",
       "               '2018-02-20 17:00:00', '2018-02-20 18:00:00',\n",
       "               '2018-02-20 19:00:00', '2018-02-20 20:00:00',\n",
       "               '2018-02-20 21:00:00', '2018-02-20 22:00:00',\n",
       "               '2018-02-20 23:00:00', '2018-02-21 00:00:00',\n",
       "               '2018-02-21 01:00:00', '2018-02-21 02:00:00',\n",
       "               '2018-02-21 03:00:00', '2018-02-21 04:00:00',\n",
       "               '2018-02-21 05:00:00', '2018-02-21 06:00:00',\n",
       "               '2018-02-21 07:00:00', '2018-02-21 08:00:00',\n",
       "               '2018-02-21 09:00:00', '2018-02-21 10:00:00',\n",
       "               '2018-02-21 11:00:00', '2018-02-21 12:00:00',\n",
       "               '2018-02-21 13:00:00', '2018-02-21 14:00:00',\n",
       "               '2018-02-21 15:00:00', '2018-02-21 16:00:00',\n",
       "               '2018-02-21 17:00:00', '2018-02-21 18:00:00',\n",
       "               '2018-02-21 19:00:00', '2018-02-21 20:00:00',\n",
       "               '2018-02-21 21:00:00', '2018-02-21 22:00:00',\n",
       "               '2018-02-21 23:00:00', '2018-02-22 00:00:00',\n",
       "               '2018-02-22 01:00:00', '2018-02-22 02:00:00',\n",
       "               '2018-02-22 03:00:00', '2018-02-22 04:00:00',\n",
       "               '2018-02-22 05:00:00', '2018-02-22 06:00:00',\n",
       "               '2018-02-22 07:00:00', '2018-02-22 08:00:00',\n",
       "               '2018-02-22 09:00:00', '2018-02-22 10:00:00',\n",
       "               '2018-02-22 11:00:00', '2018-02-22 12:00:00',\n",
       "               '2018-02-22 13:00:00', '2018-02-22 14:00:00',\n",
       "               '2018-02-22 15:00:00', '2018-02-22 16:00:00',\n",
       "               '2018-02-22 17:00:00', '2018-02-22 18:00:00',\n",
       "               '2018-02-22 19:00:00', '2018-02-22 20:00:00',\n",
       "               '2018-02-22 21:00:00', '2018-02-22 22:00:00',\n",
       "               '2018-02-22 23:00:00', '2018-02-23 00:00:00'],\n",
       "              dtype='datetime64[ns]', name='date_time', freq=None)"
      ]
     },
     "execution_count": 132,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "df =pd.read_csv('hourly_data.csv', index_col = ['date_time'], parse_dates=['date_time'])\n",
    "# df.head()\n",
    "len(df)\n",
    "\n",
    "df.iloc[-100:].index"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "beta = df.groupby([df.index.hour]).mean()\n",
    "beta = beta['available_spots']\n",
    "beta = list(beta)\n",
    "theta = df.groupby([df.index.dayofweek]).mean()\n",
    "theta = theta['available_spots']\n",
    "theta = list(theta)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "ename": "AttributeError",
     "evalue": "'numpy.ndarray' object has no attribute 'index'",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-130-33611f71809d>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0my\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m'y'\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 5\u001b[0;31m \u001b[0my\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;36m3000\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mindex\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;31mAttributeError\u001b[0m: 'numpy.ndarray' object has no attribute 'index'"
     ]
    }
   ],
   "source": [
    "data = pd.read_csv('hourly_data.csv', index_col=['date_time'], parse_dates=['date_time'])\n",
    "X = np.array(data['X'])\n",
    "y = np.array(data['y'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "empty_list = [] \n",
    "features = []\n",
    "for i in range(11):\n",
    "    features.append(list(empty_list))\n",
    "for index,row in data.iterrows():\n",
    "    features[0].append(row['X'])\n",
    "    hour = index.hour\n",
    "#     print(index)\n",
    "    day_week = index.dayofweek\n",
    "    n = int(hour/4) +1\n",
    "    features[n].append(1)\n",
    "    if day_week >4 : \n",
    "        features[10].append(1)\n",
    "        features[9].append(0)\n",
    "    else: \n",
    "        features[9].append(1)\n",
    "        features[10].append(0)\n",
    "    for i in range(1,9):\n",
    "        if i != n:\n",
    "            features[i].append(0)\n",
    "colum_names = ['X0','x1','x2','x3','x4','x5','x6','x7','x8','x9','x10']\n",
    "new_cols = np.column_stack(features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(3651, 3000, 651)"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X_train = new_cols[:3000]\n",
    "X_test = new_cols[3000:]\n",
    "len(new_cols),len(train),len(test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[  0.93135538   9.29142826   3.59417618   2.27805815 -10.39134137\n",
      "  -9.56524428   4.79292306   0.           0.           0.09363064\n",
      "  -0.09363064] 8.5080803479\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.93324173872445015"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import datasets, linear_model\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import numpy as np\n",
    "\n",
    "y_train = y[:3000]\n",
    "y_test = y[3000:]\n",
    "regr = linear_model.LinearRegression()\n",
    "regr.fit(X_train, y_train)\n",
    "print(regr.coef_,regr.intercept_)\n",
    "y_pred = regr.predict(X_test)\n",
    "regr.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfgAAAEKCAYAAAD+ckdtAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nINXOAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f7303e165f8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def rms(actual,predicted):\n",
    "    return sqrt(mean_squared_error(actual, predicted))\n",
    "\n",
    "def r2(actual,predicted):\n",
    "    mean_actual = np.mean(actual)\n",
    "    actual = np.array(actual)\n",
    "    predicted = np.array(predicted)\n",
    "    rss = np.sum((actual-predicted)**2)\n",
    "    tss = np.sum((actual-mean_actual)**2)\n",
    "    final = 1 - (rss/tss)\n",
    "    return final\n",
    "\n",
    "def accuracy_tolerance(act_set,pred_set, tolerance=10.0):\n",
    "    if len(pred_set) != len(act_set):\n",
    "        raise ValueError('prediction and actual set lengths dont match')\n",
    "    \n",
    "    n = len(pred_set)\n",
    "    count = 0\n",
    "    for i in range(n):\n",
    "        if abs(pred_set[i] - act_set[i]) <= tolerance: count += 1\n",
    "    acc = float(count) / float(n)\n",
    "    return acc\n",
    "\n",
    "def calculate_accuracy(actual, predicted,model_name='No Name'):\n",
    "    d = {}\n",
    "    d['name'] = model_name\n",
    "    d['r2'] = r2(actual,predicted)\n",
    "    d['tol'] = accuracy_tolerance(actual,predicted)\n",
    "    d['rms'] = rms(actual,predicted)\n",
    "    return d\n",
    "\n",
    "def plot_graph(y_act, y_pred):\n",
    "    # y_pred : 1D array predicted values\n",
    "    # y_act : 1D array actual values\n",
    "    if (len(y_act)!=100):\n",
    "        raise ValueError('Test set should contain 100 values only')\n",
    "    if (len(y_pred) != len(y_act)):\n",
    "        raise ValueError('actual and predicted values dont have the same lengths')\n",
    "        \n",
    "    x = [i for i in range(len(y_pred))]\n",
    "    plt.figure(figsize=(8, 4))\n",
    "    plt.plot(x, y_pred, 'r-', label='Predicted')\n",
    "    plt.plot(x, y_act, 'b-', label='Actual')\n",
    "    axes = plt.gca()\n",
    "    axes.set_ylim([0, 200])\n",
    "    plt.ylabel('available spots')\n",
    "    plt.xlabel('time')\n",
    "    plt.legend(loc='best')\n",
    "#     plt.xticks(x)\n",
    "    plt.legend()\n",
    "    plt.show()\n",
    "    \n",
    "\n",
    "plot_graph(y_pred[-100:],y_test[-100:])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD8CAYAAAB5Pm/hAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72ee933048>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "plt.plot(y[:10])\n",
    "plt.plot(y_pred[:10] , color='red')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 113,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11.554720089280151\n"
     ]
    }
   ],
   "source": [
    "from math import sqrt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "rms = sqrt(mean_squared_error(y, y_pred))\n",
    "print(rms)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 123,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.7502054231717338\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXwAAAD8CAYAAAB0IB+mAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/x/VxOHHiJdEygAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f72eeddbf28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def calc_accuracy(pred_set, act_set, tolerance=5.0):\n",
    "    if len(pred_set) != len(act_set):\n",
    "        raise ValueError('prediction and actual set lengths dont match')\n",
    "    \n",
    "    n = len(pred_set)\n",
    "    count = 0\n",
    "    dev_act = []\n",
    "    dev_pred = []\n",
    "    i_set = []\n",
    "    for i in range(n):\n",
    "        if abs(pred_set[i] - act_set[i]) <= tolerance: \n",
    "            count += 1\n",
    "        else:\n",
    "            dev_act.append(act_set[i])\n",
    "            dev_pred.append(pred_set[i])\n",
    "            i_set.append(i)\n",
    "    acc = float(count) / float(n)\n",
    "    return acc, dev_act, dev_pred, i_set\n",
    "\n",
    "acc, dev_act, dev_pred, i_set = calc_accuracy(y_pred, y, 10)\n",
    "\n",
    "print(acc)\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "\n",
    "plt.plot(np.subtract(y_pred[:100], y[:100]))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[ 0.94213464] 7.17819134841\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "1.0"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn import datasets, linear_model\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "import numpy as np\n",
    "\n",
    "X = np.array(X)\n",
    "X = np.reshape(X,(-1,1))\n",
    "\n",
    "X_train = X[:3000]\n",
    "y_train = y[:3000]\n",
    "X_test = X[3000:]\n",
    "y_test = y[3000:]\n",
    "\n",
    "\n",
    "regr = linear_model.LinearRegression()\n",
    "regr.fit(X_train, y_train)\n",
    "print(regr.coef_,regr.intercept_)\n",
    "y_pred = regr.predict(X_test)\n",
    "regr.score(X_test,y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "14.496886493118431\n"
     ]
    }
   ],
   "source": [
    "from math import sqrt\n",
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "rms = sqrt(mean_squared_error(y_test, y_pred))\n",
    "print(rms)"
   ]
  }
 ],
 "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}