bicimad-rides-data-analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# BiciMAD - Bike rides analysis\n",
    "\n",
    "Copyright © 2017 Javi Ramírez <javi.rmrz@gmail.com> | [@rameerez [tw]](http://twitter.com/rameerez) | [GitHub](http://github.com/rameerez)\n",
    "\n",
    "This code is Open Source, released under the MIT License."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Why/what\n",
    "BiciMAD has available data about their service's usage. Can we use that usage data to discover hidden patterns that can reveal underlying problems and provide BiciMad with powerful data-based suggestions that can help improve the service for all us Madrid citizens?\n",
    "\n",
    "![BiciMAD ride by sabormadrid](img/bicimad-ride.jpg)\n",
    "_Photo credit: [sabormadrid.es](http://www.sabormadrid.es/fileupload/7--bicimad-madrid-ro-623x356.jpg)_"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "## Dataset description\n",
    "\n",
    "Please refer to [EMT BiciMad OpenData docs](http://opendata.emtmadrid.es/Servicios-web/BICIMAD) for updated info. This table is only intended for illustration purposes.\n",
    "\n",
    "| Field name             | Description                                                                                |\n",
    "|------------------------|--------------------------------------------------------------------------------------------|\n",
    "| ```_id```              | Travel id                                                                                  |\n",
    "| ```user_day_code```    | User id. For a single day, all travels made by the same user will have the same user id.   |\n",
    "| ```idunplug_station``` | **Unplug**: station id.                                                                    |\n",
    "| ```idunplug_base```    | **Unplug**: base id (base within the station).                                             |\n",
    "| ```idplug_station```   | **Plug**: station id.                                                                      |\n",
    "| ```idplug_base```      | **Plug**: base id (base within the station).                                               |\n",
    "| ```unplug_hourTime```  | Hour at which the travel started, discarding info about minutes and seconds for anonymity. |\n",
    "| ```travel_time```      | Time in _seconds_ between bike unplug and bike plug back in.                               |\n",
    "| ```track```            | Travel details in GeoJSON format.                                                          |\n",
    "| ```user_type```        | 0: undetermined; 1: annual pass; 2: occasional user; 3: BiciMad employee                   |\n",
    "| ```ageRange```         | 0: undetermined; 1: [0..16]; 2: [17..18]; 3: [19..26]; 4: [27..40]; 5: [41..65] 6: [>66]   |\n",
    "| ```zip_code```         | User's postal code.                                                                        |"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Global definitions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Data describing all bike rides\n",
    "BIKE_RIDES_DATASET = './data/export_20170301_20170406.json'\n",
    "# Data describing all stations (used just for getting stations' coordinates & names)\n",
    "STATIONS_DATASET = './data/stations.json'\n",
    "\n",
    "# You need to get a Google Maps API key to plot Bokeh maps:\n",
    "# https://developers.google.com/maps/documentation/javascript/get-api-key\n",
    "GOOGLE_MAPS_API_KEY = \"AIzaSyAGMhTq1oUDAdXDgTnsVpsap7vf8LevWIw\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Imports & data loading"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import math\n",
    "import dateutil.parser\n",
    "from pandas.io.json import json_normalize"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Data visualization with [Bokeh](http://bokeh.pydata.org/en/latest/)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from bokeh.io import output_file, output_notebook, show\n",
    "from bokeh.models import (\n",
    "  GMapPlot, GMapOptions, ColumnDataSource, Circle, DataRange1d, PanTool, WheelZoomTool, BoxSelectTool\n",
    ")\n",
    "from bokeh.plotting import figure\n",
    "from bokeh.sampledata.sample_geojson import geojson"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div class=\"bk-root\">\n",
       "        <a href=\"http://bokeh.pydata.org\" target=\"_blank\" class=\"bk-logo bk-logo-small bk-logo-notebook\"></a>\n",
       "        <span id=\"7d433146-79d7-44f9-97a3-c4a306c36380\">Loading BokehJS ...</span>\n",
       "    </div>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/javascript": [
       "\n",
       "(function(global) {\n",
       "  function now() {\n",
       "    return new Date();\n",
       "  }\n",
       "\n",
       "  var force = true;\n",
       "\n",
       "  if (typeof (window._bokeh_onload_callbacks) === \"undefined\" || force === true) {\n",
       "    window._bokeh_onload_callbacks = [];\n",
       "    window._bokeh_is_loading = undefined;\n",
       "  }\n",
       "\n",
       "\n",
       "  \n",
       "  if (typeof (window._bokeh_timeout) === \"undefined\" || force === true) {\n",
       "    window._bokeh_timeout = Date.now() + 5000;\n",
       "    window._bokeh_failed_load = false;\n",
       "  }\n",
       "\n",
       "  var NB_LOAD_WARNING = {'data': {'text/html':\n",
       "     \"<div style='background-color: #fdd'>\\n\"+\n",
       "     \"<p>\\n\"+\n",
       "     \"BokehJS does not appear to have successfully loaded. If loading BokehJS from CDN, this \\n\"+\n",
       "     \"may be due to a slow or bad network connection. Possible fixes:\\n\"+\n",
       "     \"</p>\\n\"+\n",
       "     \"<ul>\\n\"+\n",
       "     \"<li>re-rerun `output_notebook()` to attempt to load from CDN again, or</li>\\n\"+\n",
       "     \"<li>use INLINE resources instead, as so:</li>\\n\"+\n",
       "     \"</ul>\\n\"+\n",
       "     \"<code>\\n\"+\n",
       "     \"from bokeh.resources import INLINE\\n\"+\n",
       "     \"output_notebook(resources=INLINE)\\n\"+\n",
       "     \"</code>\\n\"+\n",
       "     \"</div>\"}};\n",
       "\n",
       "  function display_loaded() {\n",
       "    if (window.Bokeh !== undefined) {\n",
       "      var el = document.getElementById(\"7d433146-79d7-44f9-97a3-c4a306c36380\");\n",
       "      el.textContent = \"BokehJS \" + Bokeh.version + \" successfully loaded.\";\n",
       "    } else if (Date.now() < window._bokeh_timeout) {\n",
       "      setTimeout(display_loaded, 100)\n",
       "    }\n",
       "  }\n",
       "\n",
       "  function run_callbacks() {\n",
       "    window._bokeh_onload_callbacks.forEach(function(callback) { callback() });\n",
       "    delete window._bokeh_onload_callbacks\n",
       "    console.info(\"Bokeh: all callbacks have finished\");\n",
       "  }\n",
       "\n",
       "  function load_libs(js_urls, callback) {\n",
       "    window._bokeh_onload_callbacks.push(callback);\n",
       "    if (window._bokeh_is_loading > 0) {\n",
       "      console.log(\"Bokeh: BokehJS is being loaded, scheduling callback at\", now());\n",
       "      return null;\n",
       "    }\n",
       "    if (js_urls == null || js_urls.length === 0) {\n",
       "      run_callbacks();\n",
       "      return null;\n",
       "    }\n",
       "    console.log(\"Bokeh: BokehJS not loaded, scheduling load and callback at\", now());\n",
       "    window._bokeh_is_loading = js_urls.length;\n",
       "    for (var i = 0; i < js_urls.length; i++) {\n",
       "      var url = js_urls[i];\n",
       "      var s = document.createElement('script');\n",
       "      s.src = url;\n",
       "      s.async = false;\n",
       "      s.onreadystatechange = s.onload = function() {\n",
       "        window._bokeh_is_loading--;\n",
       "        if (window._bokeh_is_loading === 0) {\n",
       "          console.log(\"Bokeh: all BokehJS libraries loaded\");\n",
       "          run_callbacks()\n",
       "        }\n",
       "      };\n",
       "      s.onerror = function() {\n",
       "        console.warn(\"failed to load library \" + url);\n",
       "      };\n",
       "      console.log(\"Bokeh: injecting script tag for BokehJS library: \", url);\n",
       "      document.getElementsByTagName(\"head\")[0].appendChild(s);\n",
       "    }\n",
       "  };var element = document.getElementById(\"7d433146-79d7-44f9-97a3-c4a306c36380\");\n",
       "  if (element == null) {\n",
       "    console.log(\"Bokeh: ERROR: autoload.js configured with elementid '7d433146-79d7-44f9-97a3-c4a306c36380' but no matching script tag was found. \")\n",
       "    return false;\n",
       "  }\n",
       "\n",
       "  var js_urls = [\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.5.min.js\", \"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.5.min.js\"];\n",
       "\n",
       "  var inline_js = [\n",
       "    function(Bokeh) {\n",
       "      Bokeh.set_log_level(\"info\");\n",
       "    },\n",
       "    \n",
       "    function(Bokeh) {\n",
       "      \n",
       "    },\n",
       "    \n",
       "    function(Bokeh) {\n",
       "      \n",
       "      document.getElementById(\"7d433146-79d7-44f9-97a3-c4a306c36380\").textContent = \"BokehJS is loading...\";\n",
       "    },\n",
       "    function(Bokeh) {\n",
       "      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-0.12.5.min.css\");\n",
       "      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-0.12.5.min.css\");\n",
       "      console.log(\"Bokeh: injecting CSS: https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.5.min.css\");\n",
       "      Bokeh.embed.inject_css(\"https://cdn.pydata.org/bokeh/release/bokeh-widgets-0.12.5.min.css\");\n",
       "    }\n",
       "  ];\n",
       "\n",
       "  function run_inline_js() {\n",
       "    \n",
       "    if ((window.Bokeh !== undefined) || (force === true)) {\n",
       "      for (var i = 0; i < inline_js.length; i++) {\n",
       "        inline_js[i](window.Bokeh);\n",
       "      }if (force === true) {\n",
       "        display_loaded();\n",
       "      }} else if (Date.now() < window._bokeh_timeout) {\n",
       "      setTimeout(run_inline_js, 100);\n",
       "    } else if (!window._bokeh_failed_load) {\n",
       "      console.log(\"Bokeh: BokehJS failed to load within specified timeout.\");\n",
       "      window._bokeh_failed_load = true;\n",
       "    } else if (force !== true) {\n",
       "      var cell = $(document.getElementById(\"7d433146-79d7-44f9-97a3-c4a306c36380\")).parents('.cell').data().cell;\n",
       "      cell.output_area.append_execute_result(NB_LOAD_WARNING)\n",
       "    }\n",
       "\n",
       "  }\n",
       "\n",
       "  if (window._bokeh_is_loading === 0) {\n",
       "    console.log(\"Bokeh: BokehJS loaded, going straight to plotting\");\n",
       "    run_inline_js();\n",
       "  } else {\n",
       "    load_libs(js_urls, function() {\n",
       "      console.log(\"Bokeh: BokehJS plotting callback run at\", now());\n",
       "      run_inline_js();\n",
       "    });\n",
       "  }\n",
       "}(this));"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "output_notebook()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Loading datasets:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [],
   "source": [
    "# Watch out! lines=True needed\n",
    "df = pd.read_json(BIKE_RIDES_DATASET, lines=True)\n",
    "\n",
    "df_stations = pd.read_json(STATIONS_DATASET)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "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>_id</th>\n",
       "      <th>ageRange</th>\n",
       "      <th>idplug_base</th>\n",
       "      <th>idplug_station</th>\n",
       "      <th>idunplug_base</th>\n",
       "      <th>idunplug_station</th>\n",
       "      <th>track</th>\n",
       "      <th>travel_time</th>\n",
       "      <th>unplug_hourTime</th>\n",
       "      <th>user_day_code</th>\n",
       "      <th>user_type</th>\n",
       "      <th>zip_code</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>{'$oid': '58d1b3dd865e9c30d0b83c6f'}</td>\n",
       "      <td>4</td>\n",
       "      <td>4</td>\n",
       "      <td>17</td>\n",
       "      <td>5</td>\n",
       "      <td>14</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2303</td>\n",
       "      <td>{'$date': '2017-03-01T00:00:00.000+0100'}</td>\n",
       "      <td>cc6d426ada96494d8dfe0fe0af512599f439cd9413d50c...</td>\n",
       "      <td>1</td>\n",
       "      <td>28012</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    _id  ageRange  idplug_base  \\\n",
       "0  {'$oid': '58d1b3dd865e9c30d0b83c6f'}         4            4   \n",
       "\n",
       "   idplug_station  idunplug_base  idunplug_station track  travel_time  \\\n",
       "0              17              5                14   NaN         2303   \n",
       "\n",
       "                             unplug_hourTime  \\\n",
       "0  {'$date': '2017-03-01T00:00:00.000+0100'}   \n",
       "\n",
       "                                       user_day_code  user_type zip_code  \n",
       "0  cc6d426ada96494d8dfe0fe0af512599f439cd9413d50c...          1    28012  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Example instance\n",
    "df[:1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Exploratory analysis\n",
    "Let's see how our data looks like"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "_id                 289981\n",
       "ageRange            289981\n",
       "idplug_base         289981\n",
       "idplug_station      289981\n",
       "idunplug_base       289981\n",
       "idunplug_station    289981\n",
       "track                87788\n",
       "travel_time         289981\n",
       "unplug_hourTime     289981\n",
       "user_day_code       289981\n",
       "user_type           289981\n",
       "zip_code            289981\n",
       "dtype: int64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see almost 290.000 bike rides recorded, but only ~87.000 of them contain GPS route information."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Travel times\n",
    "For how long do usually people ride a bike? Are there extreme values? Why?\n",
    "Do different groups of people ride bikes for different time? (for example: do tourists ride bikes for longer periods of time?)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "First of all, we're gonna need to convert travel times from seconds to minutes to better understand the measurements."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['travel_time_mins'] = df['travel_time']/60"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    289981.000000\n",
       "mean         18.274778\n",
       "std          89.918454\n",
       "min           0.000000\n",
       "25%           6.350000\n",
       "50%          10.650000\n",
       "75%          16.833333\n",
       "max        9735.966667\n",
       "Name: travel_time_mins, dtype: float64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['travel_time_mins'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Okay, people usually take bikes for around 18 minutes, which is pretty normal. The standard deviation is very high and there seems to be some extreme values (max=9735min, that's almost 7 days! – probably a stolen/lost bike? It'd be awesome to analyze that GPS track!)\n",
    "\n",
    "Let's do a log transformation to nullify the effects of extreme values and plot it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "#We need to get rid of zero values first, to avoid divide by zero errors\n",
    "df_nonzero_travel_times = df.loc[df['travel_time_mins'] > 0]\n",
    "\n",
    "df['travel_time_mins_log'] = np.log(df_nonzero_travel_times['travel_time_mins'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false,
    "scrolled": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10df31668>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYAAAAD8CAYAAAB+UHOxAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAEhSpf4f65vSg2STBOIAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a4b4940>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df['travel_time_mins_log'].hist(bins=250)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Okay, let's try to visualize the data to see our outliers. First of all, lets try to plot them all in a boxplot and see the distribution:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10e086d30>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYcAAAD9CAYAAABX0LttAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10a4b46a0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_non_extreme_travel_times = df.loc[df['travel_time_mins'] > 0]\n",
    "df_non_extreme_travel_times.boxplot(column='travel_time_mins')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Wow shit, that's some extreme values. Our data is not symmetrical at all. Look at those outliers above 6.000 minutes – that's almost 4 days using a bike! (yup, probably lost or stolen)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "15"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_non_extreme_travel_times = df.loc[df['travel_time_mins'] > 6000]\n",
    "df_non_extreme_travel_times['travel_time_mins'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So there's only 15 bikes that were used for more than 6000 minutes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "236"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_non_extreme_travel_times = df.loc[df['travel_time_mins'] > 360]\n",
    "df_non_extreme_travel_times['travel_time_mins'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And only 236 bikes that were used for more than 360 minutes (6 hours). So, let's remove extreme values and boxplot travels of less than 60 minutes (we're just \"loosing\" about 5% of the data, which are extreme outliers):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10e2b4f28>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAD9CAYAAACsq4z3AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e320d30>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_non_extreme_travel_times = df.loc[df['travel_time_mins'] < 60]\n",
    "df_non_extreme_travel_times.boxplot(column='travel_time_mins')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Okay, that's better. Now we can even see the box. As discovered before, most users take a bike for around 5-15 minutes, and we can consider outliers travels of more than 30 minutes. How many outliers are there then and what percentage of the data do they represent?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "29175"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_non_extreme_travel_times = df.loc[df['travel_time_mins'] > 30]\n",
    "df_non_extreme_travel_times['travel_time_mins'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.10061003996813585"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_non_extreme_travel_times['travel_time_mins'].count() / df['travel_time_mins'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Outliers (travels > 30min): 29175, that's about 10% of the data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Ride times under 180 seconds - number of out of order bikes?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35584"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[df['travel_time'] < 180]['_id'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.12271148799404098"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.loc[df['travel_time'] < 180]['_id'].count() / df['travel_time'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see a lot of rides under 180 seconds (3 minutes). Either Madrid have great cyclists capable of arriving to a destination within 3 minutes of taking their bikes or they take it and return it inmediatly because it was defective. This accounts for more than 12% of all rides.\n",
    "\n",
    "**1 of every 10 times you take a bike, the bike is out of order**\n",
    "\n",
    "Were they returned to the same station?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "32013"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_same_station = df.loc[(df['idunplug_station'] == df['idplug_station'])]\n",
    "df_same_station.loc[df_same_station['travel_time'] < 180]['_id'].count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Yup. Only 3000 bikes were returned in a different station in under 3 minutes (probably fast riders, or maybe there was a different station neraby)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Travel times by user type\n",
    "Do different users behave differently?\n",
    "\n",
    "Let's recall from the docs:\n",
    "```\n",
    "user_type = 0: undetermined\n",
    "user_type = 1: annual pass holder\n",
    "user_type = 2: occasional user (incl. tourists, I guess)\n",
    "user_type = 3: BiciMad employee\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>user_type</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>7375.0</td>\n",
       "      <td>30.862405</td>\n",
       "      <td>41.412250</td>\n",
       "      <td>0.033333</td>\n",
       "      <td>7.700000</td>\n",
       "      <td>15.750000</td>\n",
       "      <td>37.608333</td>\n",
       "      <td>358.766667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>247605.0</td>\n",
       "      <td>14.612580</td>\n",
       "      <td>22.656685</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>6.666667</td>\n",
       "      <td>10.416667</td>\n",
       "      <td>15.466667</td>\n",
       "      <td>359.883333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3025.0</td>\n",
       "      <td>40.599851</td>\n",
       "      <td>44.844825</td>\n",
       "      <td>0.033333</td>\n",
       "      <td>10.900000</td>\n",
       "      <td>24.166667</td>\n",
       "      <td>56.666667</td>\n",
       "      <td>349.983333</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>31962.0</td>\n",
       "      <td>41.632729</td>\n",
       "      <td>260.885889</td>\n",
       "      <td>0.016667</td>\n",
       "      <td>0.250000</td>\n",
       "      <td>16.141667</td>\n",
       "      <td>39.845833</td>\n",
       "      <td>9735.966667</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>14.0</td>\n",
       "      <td>7.164286</td>\n",
       "      <td>20.588785</td>\n",
       "      <td>0.083333</td>\n",
       "      <td>0.583333</td>\n",
       "      <td>1.008333</td>\n",
       "      <td>2.941667</td>\n",
       "      <td>78.483333</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              count       mean         std       min        25%        50%  \\\n",
       "user_type                                                                    \n",
       "0            7375.0  30.862405   41.412250  0.033333   7.700000  15.750000   \n",
       "1          247605.0  14.612580   22.656685  0.000000   6.666667  10.416667   \n",
       "2            3025.0  40.599851   44.844825  0.033333  10.900000  24.166667   \n",
       "3           31962.0  41.632729  260.885889  0.016667   0.250000  16.141667   \n",
       "4              14.0   7.164286   20.588785  0.083333   0.583333   1.008333   \n",
       "\n",
       "                 75%          max  \n",
       "user_type                          \n",
       "0          37.608333   358.766667  \n",
       "1          15.466667   359.883333  \n",
       "2          56.666667   349.983333  \n",
       "3          39.845833  9735.966667  \n",
       "4           2.941667    78.483333  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.groupby(['user_type'])['travel_time_mins'].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We see a number of things.\n",
    "\n",
    "First of all, there's a user_type = 4 that's not documented. Don't know who are those, but they account for 14 travels.\n",
    "\n",
    "Secondly, there's a **huge** difference between occasional users and regular users. While regular users take bikes for about 15 minutes (with a low std deviation), occasional users use bikes for considerably longer periods: about 40 minutes (and high 44 std dvt). This could mean occasional users are mostly tourists that use bikes for sightseeing and relaxed travels around the city. To be confirmed with GPS data.\n",
    "\n",
    "Now, look at those BiciMad employee numbers – that's where the outliers come from! It was not stolen/lost bikes, it was probably bikes under maintenance!"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Travel times by user age\n",
    "On with our next question: do users with different age behave in different ways?\n",
    "\n",
    "Let's recall from the docs:\n",
    "```\n",
    "ageRange = 0: undetermined\n",
    "ageRange = 1: from 0 to 16 years old\n",
    "ageRange = 2: from 17 to 18 years old\n",
    "ageRange = 3: from 19 to 26 years old\n",
    "ageRange = 4: from 27 to 40 years old\n",
    "ageRange = 5: from 41 to 65 years old\n",
    "ageRange = 6: older than 66\n",
    "```\n",
    "\n",
    "First of all: which of them uses BiciMAD the most? (we are going to consider only type 1 users: annual subscribers, recurrent users)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10e4f4cf8>"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY0AAAEGCAYAAACZ0MnKAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nQ0OS1M3QkCR1MzQkSd0MDUlSt/8PIN67CJR4kQ4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e4c8860>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.loc[df['user_type']==1].groupby(['ageRange'])['_id'].count().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We have a winner! People aged 27-40 use BiciMAD the most. For how long on average?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10e538be0>"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAAEKCAYAAAACS67iAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nRCKEAl1EJEIo0EVEIsT/Bxw4XGfwcsiIAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10e564ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.loc[df['user_type']==1].groupby(['ageRange'])['travel_time_mins'].mean().plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Not a lot of differences. Younger and older users seem to use bikes for longer periods, while the vast majority of users (27-40 yo) use them for the least amount of time. This may be because they ride faster, or because they use BiciMAD for functional purposes (commuting / quickly cruising through the city), as opposed to using it for recreational purposes / more relaxed use.\n",
    "\n",
    "There is only a 6 minute difference though."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Popular hours\n",
    "\n",
    "to-do: json_normalize column unplug_hourTime"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Popular hours by station per day\n",
    "\n",
    "to-do"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### User behavior - rides per user per day\n",
    "\n",
    "How many times per day do users use BiciMAD? Let's just consider type 1 users (annual subscribers) for now:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.loc[df['user_type']==1].groupby(['user_day_code'])['_id'].count().describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hold on. Did someone really take 27 bikes in a single day?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.loc[df['user_type']==1].groupby(['user_day_code'])['_id'].count().sort_values(ascending=False)[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.loc[df['user_day_code']=='9706face97e41bf2c2f9fb84e606628c7755b62c65f9e3d15971d8abe4daf5a2']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What the fuck. He spent almost 3 hours (from ~9 to ~11) undocking bikes and docking them again within seconds. In 3-4 close stations. One was undocked for 5+ hours, probably he didn't dock it back correctly."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Is there any difference with type 2 (occasional users)?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.loc[df['user_type']==2].groupby(['user_day_code'])['_id'].count().describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Nah, both types of users take 1-2 bikes per day."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Demographics - ZIP code"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df.groupby(['zip_code'])['_id'].count()[:5] # showing just the 5 first instances"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Wow, that's a lot of shit messing up with our nice ZIP codes. Email addresses? Seriously? Gawd, let's try getting ride of those non-numeric values first."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def make_int(s):\n",
    "    s = s.strip()\n",
    "    return int(s) if (s and s.isdigit()) else 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['zip_code_as_int'] = df['zip_code'].apply(lambda x: make_int(x))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Ah, that's better."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "# to-do: deal with this stack overflow\n",
    "df.groupby(['zip_code_as_int'])['_id'].count()[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Most popular stations\n",
    "Which are the most popular stations? Which are mainly used for taking bikes and which are mainly used to return bikes? Are there any popular travels between 2 stations?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['idunplug_station'].value_counts().sort_values(ascending=False)[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df['idplug_station'].value_counts().sort_values(ascending=False)[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Well, from this we can see that the most used stations are the most used in _both_ ways: for providing bikes as well as for receiving bikes. There is no asymmetry, no stations that are mainly \"bike providers\" and others that are mainly \"bike receivers\". This makes sense: if a station is placed in a good spot and it's well known among users, they will use that station for convenience and for lack of knowing alternatives, regardless whether they wanted to take a bike or to leave one."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# to-do: search for most popular pairs of stations (most popular rides)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Stations and maps\n",
    "\n",
    "Let's add some info about stations and their GPS location to our rides dataset.\n",
    "\n",
    "⚠️ to-do: cleanup this mess"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df2 = pd.merge(df, df_stations, left_on='idunplug_station', right_on='id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations[:1] # not displaying the full dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "s_rides_departing_from = df2.groupby('id')['_id'].count()\n",
    "df_rides_departing_from = pd.DataFrame({'id':s_rides_departing_from.index, 'rides_departing_from': s_rides_departing_from.values})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations = pd.merge(df_rides_departing_from, df_stations, left_on='id', right_on='id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "df_stations['departing_point_size'] = df_stations['rides_departing_from'] / df2['_id'].count()*1000\n",
    "df_stations[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df3 = pd.merge(df, df_stations, left_on='idplug_station', right_on='id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "s_rides_arriving_to = df3.groupby('id')['_id'].count()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_rides_arriving_to = pd.DataFrame({'id':s_rides_arriving_to.index, 'rides_arriving_to': s_rides_arriving_to.values})\n",
    "df_rides_arriving_to[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations = pd.merge(df_rides_arriving_to, df_stations, left_on='id', right_on='id')\n",
    "df_stations[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations['arriving_point_size'] = df_stations['rides_arriving_to'] / df3['_id'].count()*1000\n",
    "df_stations[:1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations.sort('rides_arriving_to')[:3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_stations.sort('rides_departing_from', ascending=False)[:100]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**the most unused stations are just in the center of Madrid!**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "map_options = GMapOptions(lat=40.42, lng=-3.70, map_type=\"roadmap\", zoom=11)\n",
    "\n",
    "plot = GMapPlot(\n",
    "    x_range=DataRange1d(), y_range=DataRange1d(), map_options=map_options\n",
    ")\n",
    "plot.title.text = \"BiciMAD stations - stations with most PLUGGED bikes\"\n",
    "\n",
    "# For GMaps to function, Google requires you obtain and enable an API key:\n",
    "plot.api_key = GOOGLE_MAPS_API_KEY\n",
    "\n",
    "source = ColumnDataSource(df_stations)\n",
    "\n",
    "circle = Circle(x=\"longitude\", y=\"latitude\", size=\"arriving_point_size\", fill_color=\"blue\", fill_alpha=0.8, line_color=None)\n",
    "plot.add_glyph(source, circle)\n",
    "\n",
    "plot.add_tools(PanTool(), WheelZoomTool(), BoxSelectTool())\n",
    "output_file(\"gmap_plot.html\")\n",
    "show(plot)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Most profitable stations\n",
    "\n",
    "According to the [Pareto principle](https://en.wikipedia.org/wiki/Pareto_principle), 80% of BiciMAD's revenue should come from the top 20 stations. We know which are the most popular stations, but do those make the most amount of revenue? We have to compute the revenue per ride first, and then gruop all revenues by stations.\n",
    "\n",
    "From [BiciMAD's official pricing website](https://www.bicimad.com/index.php?s=tarifas):\n",
    "\n",
    "_(I'm not considering fixed costs such as the annual pass renewal cost, just ride-dependent costs, so this won't reflect BiciMAD **total** revenue, just ride-related revenue. Also, bonuses for taking bikes from high-occupation stations and so are not yet considered)_\n",
    "\n",
    "| Annual Member                                               | Occasional User                                             |\n",
    "|-------------------------------------------------------------|-------------------------------------------------------------|\n",
    "| First 30 minutes: **0,50€**                                 | First hour: **2€**                                          |\n",
    "| Next 30 minutes: **0,60€**                                  | Second hour: **4€**                                         |\n",
    "| Penalty for exceeding 2 hours, each additional hour: **4€** | Penalty for exceeding 2 hours, each additional hour: **4€** |\n",
    "| Bonuses not considered (yet)                                | Bonuses not considered (yet)                                |"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "USER_TYPE_ANNUAL = 1\n",
    "USER_TYPE_OCCASIONAL = 2\n",
    "\n",
    "def ride_to_revenue(user_type, ride_time):\n",
    "    revenue = 0\n",
    "    if user_type == USER_TYPE_ANNUAL:\n",
    "        if ride_time <= 1800: # less than 30 minutes\n",
    "            revenue = 0.5\n",
    "        elif ride_time > 1800:\n",
    "            revenue = 0.5 + ( 0.6 * math.ceil((ride_time-1800)/1800) )\n",
    "        if ride_time >= 7200:  # more than 2 hours\n",
    "            revenue += 4 * math.ceil((ride_time-7200)/3600) # how many hours are exceeding 2h\n",
    "            \n",
    "    elif user_type == USER_TYPE_OCCASIONAL:\n",
    "        if ride_time <= 3600: # less than 1 hour\n",
    "            revenue = 2\n",
    "        elif ride_time > 3600: # between 1 hour and 2 hours\n",
    "            revenue = 2 + 4\n",
    "        if ride_time > 7200:\n",
    "            revenue += 4 * math.ceil((ride_time-7200)/3600) # how many hours are exceeding 2h\n",
    "    \n",
    "    else:\n",
    "        revenue = 0\n",
    "    return revenue   "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def compute_revenue(row):\n",
    "    return ride_to_revenue(row['user_type'], row['travel_time'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "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>idunplug_station</th>\n",
       "      <th>idplug_station</th>\n",
       "      <th>travel_time_mins</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14</td>\n",
       "      <td>17</td>\n",
       "      <td>38.383333</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>123</td>\n",
       "      <td>76</td>\n",
       "      <td>14.600000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>166</td>\n",
       "      <td>163</td>\n",
       "      <td>8.050000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>113</td>\n",
       "      <td>69</td>\n",
       "      <td>1.750000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>10</td>\n",
       "      <td>63</td>\n",
       "      <td>5.450000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>4</td>\n",
       "      <td>17</td>\n",
       "      <td>8.000000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>142</td>\n",
       "      <td>71</td>\n",
       "      <td>17.233333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>74</td>\n",
       "      <td>80</td>\n",
       "      <td>4.383333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>30</td>\n",
       "      <td>22</td>\n",
       "      <td>2.500000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>55</td>\n",
       "      <td>129</td>\n",
       "      <td>2.300000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>26</td>\n",
       "      <td>38</td>\n",
       "      <td>3.950000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>146</td>\n",
       "      <td>118</td>\n",
       "      <td>9.600000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>130</td>\n",
       "      <td>17</td>\n",
       "      <td>6.783333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>168</td>\n",
       "      <td>168</td>\n",
       "      <td>15.466667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>59</td>\n",
       "      <td>57</td>\n",
       "      <td>9.083333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>23</td>\n",
       "      <td>23</td>\n",
       "      <td>0.300000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>30.366667</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>31.766667</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>31.900000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>149</td>\n",
       "      <td>30</td>\n",
       "      <td>14.333333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>45</td>\n",
       "      <td>128</td>\n",
       "      <td>4.633333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>36.133333</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>36.800000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>14</td>\n",
       "      <td>25</td>\n",
       "      <td>34.950000</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>62</td>\n",
       "      <td>48</td>\n",
       "      <td>21.983333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>46</td>\n",
       "      <td>128</td>\n",
       "      <td>6.133333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>21</td>\n",
       "      <td>129</td>\n",
       "      <td>7.150000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>107</td>\n",
       "      <td>157</td>\n",
       "      <td>10.100000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>52</td>\n",
       "      <td>10</td>\n",
       "      <td>6.183333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>44</td>\n",
       "      <td>174</td>\n",
       "      <td>24.400000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289951</th>\n",
       "      <td>46</td>\n",
       "      <td>129</td>\n",
       "      <td>5.133333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289952</th>\n",
       "      <td>57</td>\n",
       "      <td>43</td>\n",
       "      <td>5.633333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289953</th>\n",
       "      <td>27</td>\n",
       "      <td>71</td>\n",
       "      <td>7.316667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289954</th>\n",
       "      <td>43</td>\n",
       "      <td>55</td>\n",
       "      <td>5.300000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289955</th>\n",
       "      <td>38</td>\n",
       "      <td>27</td>\n",
       "      <td>3.733333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289956</th>\n",
       "      <td>2</td>\n",
       "      <td>47</td>\n",
       "      <td>5.500000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289957</th>\n",
       "      <td>125</td>\n",
       "      <td>38</td>\n",
       "      <td>13.183333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289958</th>\n",
       "      <td>146</td>\n",
       "      <td>172</td>\n",
       "      <td>5.166667</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289959</th>\n",
       "      <td>96</td>\n",
       "      <td>166</td>\n",
       "      <td>6.833333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289960</th>\n",
       "      <td>42</td>\n",
       "      <td>42</td>\n",
       "      <td>0.450000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289961</th>\n",
       "      <td>86</td>\n",
       "      <td>82</td>\n",
       "      <td>7.500000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289962</th>\n",
       "      <td>142</td>\n",
       "      <td>156</td>\n",
       "      <td>6.700000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289963</th>\n",
       "      <td>57</td>\n",
       "      <td>174</td>\n",
       "      <td>26.183333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289964</th>\n",
       "      <td>45</td>\n",
       "      <td>79</td>\n",
       "      <td>15.450000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289965</th>\n",
       "      <td>57</td>\n",
       "      <td>174</td>\n",
       "      <td>25.650000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289966</th>\n",
       "      <td>45</td>\n",
       "      <td>79</td>\n",
       "      <td>15.383333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289967</th>\n",
       "      <td>5</td>\n",
       "      <td>23</td>\n",
       "      <td>5.216667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289968</th>\n",
       "      <td>129</td>\n",
       "      <td>132</td>\n",
       "      <td>18.400000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289969</th>\n",
       "      <td>157</td>\n",
       "      <td>157</td>\n",
       "      <td>0.233333</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289970</th>\n",
       "      <td>84</td>\n",
       "      <td>116</td>\n",
       "      <td>17.466667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289971</th>\n",
       "      <td>47</td>\n",
       "      <td>43</td>\n",
       "      <td>4.500000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289972</th>\n",
       "      <td>12</td>\n",
       "      <td>142</td>\n",
       "      <td>21.100000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289973</th>\n",
       "      <td>117</td>\n",
       "      <td>16</td>\n",
       "      <td>6.416667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289974</th>\n",
       "      <td>59</td>\n",
       "      <td>135</td>\n",
       "      <td>21.050000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289975</th>\n",
       "      <td>38</td>\n",
       "      <td>38</td>\n",
       "      <td>0.216667</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289976</th>\n",
       "      <td>26</td>\n",
       "      <td>45</td>\n",
       "      <td>7.200000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289977</th>\n",
       "      <td>121</td>\n",
       "      <td>35</td>\n",
       "      <td>5.783333</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289978</th>\n",
       "      <td>43</td>\n",
       "      <td>12</td>\n",
       "      <td>10.800000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289979</th>\n",
       "      <td>54</td>\n",
       "      <td>133</td>\n",
       "      <td>5.650000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>289980</th>\n",
       "      <td>90</td>\n",
       "      <td>32</td>\n",
       "      <td>6.250000</td>\n",
       "      <td>0.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>289981 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        idunplug_station  idplug_station  travel_time_mins  revenue\n",
       "0                     14              17         38.383333      1.1\n",
       "1                    123              76         14.600000      0.5\n",
       "2                    166             163          8.050000      0.5\n",
       "3                    113              69          1.750000      0.5\n",
       "4                     10              63          5.450000      0.5\n",
       "5                      4              17          8.000000      0.0\n",
       "6                    142              71         17.233333      0.5\n",
       "7                     74              80          4.383333      0.5\n",
       "8                     30              22          2.500000      0.5\n",
       "9                     55             129          2.300000      0.5\n",
       "10                    26              38          3.950000      0.5\n",
       "11                   146             118          9.600000      0.5\n",
       "12                   130              17          6.783333      0.5\n",
       "13                   168             168         15.466667      0.5\n",
       "14                    59              57          9.083333      0.5\n",
       "15                    23              23          0.300000      0.5\n",
       "16                    14              25         30.366667      0.0\n",
       "17                    14              25         31.766667      0.0\n",
       "18                    14              25         31.900000      0.0\n",
       "19                   149              30         14.333333      0.5\n",
       "20                    45             128          4.633333      0.5\n",
       "21                    14              25         36.133333      0.0\n",
       "22                    14              25         36.800000      0.0\n",
       "23                    14              25         34.950000      0.0\n",
       "24                    62              48         21.983333      0.5\n",
       "25                    46             128          6.133333      0.5\n",
       "26                    21             129          7.150000      0.5\n",
       "27                   107             157         10.100000      0.5\n",
       "28                    52              10          6.183333      0.5\n",
       "29                    44             174         24.400000      0.5\n",
       "...                  ...             ...               ...      ...\n",
       "289951                46             129          5.133333      0.5\n",
       "289952                57              43          5.633333      0.5\n",
       "289953                27              71          7.316667      0.5\n",
       "289954                43              55          5.300000      0.5\n",
       "289955                38              27          3.733333      0.5\n",
       "289956                 2              47          5.500000      0.5\n",
       "289957               125              38         13.183333      0.5\n",
       "289958               146             172          5.166667      0.0\n",
       "289959                96             166          6.833333      0.5\n",
       "289960                42              42          0.450000      0.5\n",
       "289961                86              82          7.500000      0.5\n",
       "289962               142             156          6.700000      0.5\n",
       "289963                57             174         26.183333      0.5\n",
       "289964                45              79         15.450000      0.5\n",
       "289965                57             174         25.650000      0.5\n",
       "289966                45              79         15.383333      0.5\n",
       "289967                 5              23          5.216667      0.5\n",
       "289968               129             132         18.400000      0.5\n",
       "289969               157             157          0.233333      0.0\n",
       "289970                84             116         17.466667      0.5\n",
       "289971                47              43          4.500000      0.5\n",
       "289972                12             142         21.100000      0.5\n",
       "289973               117              16          6.416667      0.5\n",
       "289974                59             135         21.050000      0.5\n",
       "289975                38              38          0.216667      0.5\n",
       "289976                26              45          7.200000      0.5\n",
       "289977               121              35          5.783333      0.5\n",
       "289978                43              12         10.800000      0.5\n",
       "289979                54             133          5.650000      0.5\n",
       "289980                90              32          6.250000      0.5\n",
       "\n",
       "[289981 rows x 4 columns]"
      ]
     },
     "execution_count": 76,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['revenue'] = df.apply(compute_revenue, axis=1)\n",
    "df[['idunplug_station', 'idplug_station', 'travel_time_mins', 'revenue']]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cool. But since most long rides belong to maintenance, let's filter by user type (`user_type` 1 is an annual pass holder). And lets filter extreme rides as we saw earlier, keeping just rides of less than 60 min, which still accounts for more than 90% of the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {
    "collapsed": false
   },
   "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>idunplug_station</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>14</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   idunplug_station  revenue\n",
       "0                14      1.1"
      ]
     },
     "execution_count": 77,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_revenue = df[(df['user_type'] == 1) & (df['travel_time_mins'] < 60)][['idunplug_station', 'revenue']].sort_values(by=['revenue'], ascending=False)\n",
    "df_revenue[:1]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cool! Now let's join some info about each station to get a more human-readable table."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "df_revenue = df_revenue.join(df_stations.set_index('id'), on=\"idunplug_station\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {
    "collapsed": false
   },
   "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>name</th>\n",
       "      <th>revenue</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Conde Duque</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>83944</th>\n",
       "      <td>San Andrés</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>227211</th>\n",
       "      <td>General Martínez Campos</td>\n",
       "      <td>1.1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                            name  revenue\n",
       "0                    Conde Duque      1.1\n",
       "83944                 San Andrés      1.1\n",
       "227211  General Martínez Campos       1.1"
      ]
     },
     "execution_count": 79,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_revenue[['name','revenue']][:3]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Cool. Note that the attribution of the revenue goes to the station from which the bike was unplugged.\n",
    "\n",
    "Let's group by stations and sum the revenue to obtain a total revenue per station. And let's sort the result, too."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 93,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "name\n",
       "Plaza de Lavapiés                      1848.6\n",
       "Plaza de la Cebada                     1711.6\n",
       "General Álvarez de Castro              1675.2\n",
       "Santa María de la Cabeza               1517.7\n",
       "Puerta de Madrid                       1452.2\n",
       "Antón Martín                           1362.8\n",
       "Plaza de la Independencia              1351.5\n",
       "Cea Bermúdez                           1347.1\n",
       "Glorieta de los Cuatro Caminos         1280.6\n",
       "Fernando el Católico                   1248.0\n",
       "Ortega y Gasset                        1221.1\n",
       "Entrada Matadero                       1220.3\n",
       "Velázquez 130                          1212.1\n",
       "Barceló                                1190.2\n",
       "Red de San Luis A                      1186.0\n",
       "Alonso Martínez                        1161.1\n",
       "José Abascal                           1158.4\n",
       "Palos de la Frontera                   1158.0\n",
       "Puerta del Sol A                       1154.4\n",
       "Puerta de Toledo                       1140.6\n",
       "Puerta de Granada                      1119.0\n",
       "Juan Bravo 50                          1109.6\n",
       "Jacinto Benavente                      1070.3\n",
       "Puerta del Pacífico                    1069.7\n",
       "Puerta del Doce de Octubre             1060.0\n",
       "Diego de León 52                       1057.5\n",
       "Plaza de San Ildefonso                 1044.8\n",
       "San Hermenegildo                       1035.7\n",
       "Menéndez Pelayo                         991.3\n",
       "Bárbara de Braganza                     979.5\n",
       "                                        ...  \n",
       "Méndez Álvaro                           454.3\n",
       "José Gutiérrez Abascal                  449.5\n",
       "Alcántara                               431.0\n",
       "Palacio de Deportes                     428.1\n",
       "Plaza de la República Argentina         423.9\n",
       "Ferraz                                  421.9\n",
       "Orense 36                               416.5\n",
       "Castelló                                415.7\n",
       "Atocha B                                415.0\n",
       "Santa Cruz del Marcenado                414.6\n",
       "Claudio Coello                          404.1\n",
       "Alberto Alcocer                         392.0\n",
       "Castellana 106                          390.1\n",
       "Velázquez                               382.1\n",
       "Plaza de Colón                          381.4\n",
       "Colón A                                 381.2\n",
       "Plaza de la Provincia                   371.2\n",
       "Banco de España B                       352.6\n",
       "Castellana frente a Hermanos Pinzón     351.5\n",
       "Serrano 210                             339.1\n",
       "Pintor Rosales                          320.8\n",
       "Paseo de la Habana 63                   302.0\n",
       "Red de San Luis B                       301.4\n",
       "Pío Baroja                              300.3\n",
       "Carretas                                288.3\n",
       "Raimundo Fernández                      275.5\n",
       "Plaza de Celenque B                     265.6\n",
       "Plaza del Carmen                        227.8\n",
       "Plaza de Celenque A                     192.5\n",
       "Paseo de Moret                          183.1\n",
       "Name: revenue, Length: 172, dtype: float64"
      ]
     },
     "execution_count": 93,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stations_with_revenue = df_revenue.groupby(['name']).agg('sum').sort_values(by=['revenue'], ascending=False)['revenue']\n",
    "stations_with_revenue"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What's the total revenue?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 94,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "125904.5"
      ]
     },
     "execution_count": 94,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total_revenue = stations_with_revenue.agg('sum')\n",
    "total_revenue"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Do the top 20 stations account for 80% of the revenue?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 103,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.21125138497829704"
      ]
     },
     "execution_count": 103,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_revenue.groupby(['name']).agg('sum').sort_values(by=['revenue'], ascending=False)['revenue'][:20].agg('sum') / total_revenue"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Hmmm... no, not really."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x10ec24e48>"
      ]
     },
     "execution_count": 99,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYEAAAHzCAYAAADGs6HuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ng0BGRkbGYoz/DwDpLcqd2NgSAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x10ed5fcf8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "stations_with_revenue.plot(kind=\"bar\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How many stations below the mean?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 108,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "732.0029069767442"
      ]
     },
     "execution_count": 108,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stations_with_revenue.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "98"
      ]
     },
     "execution_count": 120,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stations_with_revenue.where(stations_with_revenue < stations_with_revenue.mean()).count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And that accounts for... % of the total stations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.56976744186046513"
      ]
     },
     "execution_count": 121,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "stations_with_revenue.where(stations_with_revenue < stations_with_revenue.mean()).count() / stations_with_revenue.count()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "So almost 60% of the stations generate less revenue than the mean."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Bias warning**\n",
    "\n",
    "- Not considering trips for users other than annual pass holders\n",
    "- Not considering trips of more than 1 hour\n",
    "- Data does not belong to a single month, that's not revenue per station per month"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# ⚠️ Work in progress! ⚠️\n",
    "\n",
    "## To-do:\n",
    "\n",
    "- Compare different stations along time\n",
    "- Are there oversized stations?\n",
    "- Any bike that was unplugged but not plugged in back again (stolen bikes)?\n",
    "- Usage: functional or recreative?\n",
    "- Where do BiciMAD's users live? (postal code?)\n",
    "- Are there round trips? i.e.: same daily user code and trips with same origin and destination just different order\n",
    "- Machine learning: when will a bike fail / need maintenance?\n",
    "- Machine learning: will there be more bike trips when the weather is better?"
   ]
  }
 ],
 "metadata": {
  "celltoolbar": "Raw Cell Format",
  "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": "3.5.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}