challenge_2016_hourly_average.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Average number of bikes per hour and per day\n",
    "\n",
    "This notebooks calculates the average number of bikes per hour and per day for each station to be predicted. It then predicts the number of expected bikes for the test set by using a nested lookup table where the keys are the days and the hours. This method scored 3.947."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "\n",
    "def get_station_updates(city, station):\n",
    "    df = pd.read_csv('challenge_data/part_1/{}/stations/{}.csv'.format(city, station))\n",
    "    df['moment'] = pd.to_datetime(df['moment'])\n",
    "    return df"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fit"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def mean_bikes_per_day_per_hour(city, station):\n",
    "    df = get_station_updates(city, station)\n",
    "    df['weekday'] = df['moment'].map(lambda x: x.weekday())\n",
    "    df['hour'] = df['moment'].map(lambda x: x.hour)\n",
    "    means = {\n",
    "        weekday: {\n",
    "            hour: group2['bikes'].mean()\n",
    "            for hour, group2 in group1.groupby('hour')\n",
    "        }\n",
    "        for weekday, group1 in df.groupby('weekday')\n",
    "    }\n",
    "    return means"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The lookup table can be visualized with a heatmap. In the following heatmap, each row indicates an hour and each column represents a day. The cells are \"warmer\" the more bikes there during that hour and day."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.text.Text at 0x109670668>"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAi4AAAF3CAYAAACCI8B9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nf9Kpq/zF2l7TAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x106cb4d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import calendar\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "%matplotlib inline\n",
    "\n",
    "lookup_table = mean_bikes_per_day_per_hour('toulouse', '00003-pomme')\n",
    "hm = pd.DataFrame.from_dict(lookup_table, orient='index').T\n",
    "hm.columns = [calendar.day_name[col] for col in hm.columns]\n",
    "\n",
    "plt.figure(figsize=(10, 6))\n",
    "sns.heatmap(hm, linewidths=.4, cmap='YlOrBr')\n",
    "plt.xlabel('Days', fontsize=16)\n",
    "plt.ylabel('Hours', fontsize=16)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Predict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "to_predict_df = pd.read_csv('challenge_data/part_1/test-blank.csv', index_col=0)\n",
    "to_predict_df['moment'] = pd.to_datetime(to_predict_df.index)\n",
    "to_predict_df['weekday'] = to_predict_df['moment'].map(lambda t: t.weekday())\n",
    "to_predict_df['hour'] = to_predict_df['moment'].map(lambda t: t.hour)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Build a dictionary with the name of the cities and the stations to predict."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "means = {\n",
    "    city_name: {\n",
    "        station_name: mean_bikes_per_day_per_hour(city_name, station_name)\n",
    "        for station_name in rows['station'].unique()\n",
    "    }\n",
    "    for city_name, rows in to_predict_df.groupby('city')\n",
    "}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For each row in the test set, get the average number of bikes from the lookup table."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "to_predict_df['bikes'] = to_predict_df.apply(lambda r: means[r['city']][r['station']][r['weekday']][r['hour']], axis=1).tolist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Delete the columns that are not necessary for submitting."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "to_predict_df.drop('moment', axis=1, inplace=True)\n",
    "to_predict_df.drop('weekday', axis=1, inplace=True)\n",
    "to_predict_df.drop('hour', axis=1, inplace=True)"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}