MAUP.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "MAUP",
      "provenance": [],
      "collapsed_sections": [
        "62wjhPqcZKND",
        "JQdLhQWyBwU_",
        "oUF-nyuBB3EQ",
        "Iy-WNCWGNHmL",
        "Dws_lK9ZJgtX",
        "JJw-u8b4JQ2Y",
        "rlBZR9t3D4zc",
        "_s4ZljhRERhw",
        "O6Ku5OKEFeW9",
        "tOJKfEga2B7n"
      ],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JWTpmEbKBbVu",
        "colab_type": "text"
      },
      "source": [
        "### Setup and Initialisation"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "62wjhPqcZKND",
        "colab_type": "text"
      },
      "source": [
        "##### Installing Rtree [Only for Google Colab]"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "efUdAUzYZS9j",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import os\n",
        "!apt-get install -qq curl g++ make\n",
        "!curl -L http://download.osgeo.org/libspatialindex/spatialindex-src-1.8.5.tar.gz | tar xz\n",
        "os.chdir('spatialindex-src-1.8.5')\n",
        "!./configure\n",
        "!make\n",
        "!make install\n",
        "!ldconfig\n",
        "!pip install rtree\n",
        "\n",
        "#restart runtime\n",
        "os.kill(os.getpid(), 9)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JQdLhQWyBwU_",
        "colab_type": "text"
      },
      "source": [
        "##### Library Installs"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6gZwFhH0BTVf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!pip install h3\n",
        "!pip install geopandas\n",
        "!pip install hdbscan"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oUF-nyuBB3EQ",
        "colab_type": "text"
      },
      "source": [
        "##### Imports"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EaoZ2BcMBgVI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import warnings\n",
        "import os\n",
        "from datetime import datetime\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import geopandas as gpd\n",
        "from geopandas.tools import sjoin\n",
        "from shapely.geometry import Point\n",
        "from h3 import h3\n",
        "\n",
        "\n",
        "warnings.filterwarnings('ignore')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Iy-WNCWGNHmL",
        "colab_type": "text"
      },
      "source": [
        "### Auxiliary Functions"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EtN7S89mNLm2",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def geoToHex(data, res):\n",
        "  '''\n",
        "  Computes Hexagon IDs from Latitude, Longitude data. For more information visit: https://github.com/uber/h3-py\n",
        "\n",
        "  Input: data: 2D array of geolocations [[lat, lng],]\n",
        "         res: resolution at which the hexagon should be computed.\n",
        "  Output: 1D array of hexagon ids [hex_id,]\n",
        "  '''\n",
        "  l = []\n",
        "  for i in data:\n",
        "    l.append(h3.geo_to_h3(i[0], i[1], res))\n",
        "  return l\n",
        "\n",
        "\n",
        "\n",
        "def computeMPPerHexagon(data, key, writeCsv=True):\n",
        "  '''\n",
        "  This function computes Market potential M.P for an areal unit X\n",
        "  Input: data->Pandas dataframe, key-> column for unique identifier for the area X (a hexagon)\n",
        "  Output: Pandas DataFrame: [hex_id, market_potential, total_ubertaxi_rides, total_yellowtaxi_rides] and/or csv for the same\n",
        "  \n",
        "  '''\n",
        "\n",
        "  #one-hot encde taxi type\n",
        "  x = pd.get_dummies(data['type'])\n",
        "\n",
        "  #concatenate with data\n",
        "  x = pd.concat([data,pd.get_dummies(data['type'])],axis=1)\n",
        "\n",
        "  #group by selected area : hexagon\n",
        "  x = x.groupby(key).agg({\"uber_taxi\":\"sum\", \"yellow_taxi\":\"sum\"}).reset_index()\n",
        "\n",
        "  #find market pootential\n",
        "  MP = x['yellow_taxi'] / ( x['yellow_taxi'] + x['uber_taxi'] )\n",
        "  x['market_potential'] = MP\n",
        "\n",
        "  #rename columns to their new meanings\n",
        "  x.rename(columns={key:\"h3_id\", \"uber_taxi\":\"total_ubertaxi_rides\", \"yellow_taxi\":\"total_yellowtaxi_rides\"}, inplace=True)\n",
        " \n",
        "  #generate csv for mapping on keplers\n",
        "  if(writeCsv):\n",
        "    x.to_csv(\"results/\"+key+\".csv\", index=False)\n",
        "\n",
        "  return x\n",
        "\n",
        "\n",
        "\n",
        "def computeMPPerPolygon(data, key, writeCsv=True):\n",
        "  '''\n",
        "  This function computes Market potential M.P for an areal unit X\n",
        "  Input: data->Pandas dataframe, key-> column for unique identifier for the area X (a polygon), area_geometry-> column for gemetry for area defined by key\n",
        "  Output: Pandas DataFrame: [area_id, area_geometry, market_potential, total_ubertaxi_rides, total_yellowtaxi_rides] and/or csv for the same\n",
        "  \n",
        "  '''\n",
        "  #one-hot encde taxi type\n",
        "  x = pd.get_dummies(data['type'])\n",
        "\n",
        "  #concatenate with data\n",
        "  x = pd.concat([data,pd.get_dummies(data['type'])],axis=1)\n",
        "\n",
        "  #parse to int datatype\n",
        "  x['yellow_taxi'] = x['yellow_taxi'].astype(int)\n",
        "  x['uber_taxi'] = x['uber_taxi'].astype(int)\n",
        "\n",
        "  #group by selected area : hexagon\n",
        "  x = x.groupby(key+\"_id\").agg({\"uber_taxi\":\"sum\", \"yellow_taxi\":\"sum\", key+\"_geometry\": list}).reset_index()\n",
        "\n",
        "  #extract gemetries from sets.\n",
        "  x[key+\"_geometry\"] = x[key+\"_geometry\"].apply(lambda x: x[0])\n",
        "\n",
        "  #find market pootential\n",
        "  MP = x['yellow_taxi'] / ( x['yellow_taxi'] + x['uber_taxi'] )\n",
        "  x['market_potential'] = MP\n",
        "\n",
        "  #rename columns to their new meanings\n",
        "  x.rename(columns={key:\"h3_id\", \"uber_taxi\":\"total_ubertaxi_rides\", \"yellow_taxi\":\"total_yellowtaxi_rides\"}, inplace=True)\n",
        "  \n",
        "  #generate csv for mapping on keplers\n",
        "  if(writeCsv):\n",
        "    x.to_csv(\"results\"+\"/\"+key+\".csv\", index=False)\n",
        "\n",
        "  return x\n",
        "\n",
        "\n",
        "def cutoff(data, cutoff):\n",
        "  ''' \n",
        "  This function filters a pandas dataframe by a given cutoff (an integer)\n",
        "  '''\n",
        "  data = data[ (data.total_ubertaxi_rides >= cutoff) & (data.total_yellowtaxi_rides>=cutoff)]\n",
        "  return data\n",
        "\n",
        "\n",
        "\n",
        "def getPolygon(x):\n",
        "  ''' This function computes the convex hull of a set of latitude, longitude.\n",
        "  Input: 2D numpy array of floats\n",
        "  Output: NaN if convexhull cannot be resolved\n",
        "          Shapely POLYGON geometry (the convex hull)\n",
        "  '''\n",
        "  x = x[['lon', 'lat']].values\n",
        "  x = np.unique(x, axis=0)\n",
        "  if(x.shape[0]<=2):\n",
        "    return np.nan\n",
        "  try:\n",
        "    hull = ConvexHull(x)\n",
        "    return Polygon(x[hull.vertices])\n",
        "  except:\n",
        "    return np.nan\n",
        "\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KJl55yyjJdST",
        "colab_type": "text"
      },
      "source": [
        "### Main Code"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dttG6CjgFH68",
        "colab_type": "text"
      },
      "source": [
        "#### Optional Section"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Dws_lK9ZJgtX",
        "colab_type": "text"
      },
      "source": [
        "##### Process Raw Yellow Taxi Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P1PeCjSOWnMC",
        "colab_type": "code",
        "outputId": "cc8fe764-6d64-4ea8-9135-fcdcdd3f3402",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "from google.colab import auth\n",
        "import google.datalab.bigquery as bq\n",
        "\n",
        "\n",
        "auth.authenticate_user()\n",
        "YOUR_DATALAB_PROJECT_ID = '' #enter this\n",
        "os.environ['PROJECT_ID'] = YOUR_DATALAB_PROJECT_ID\n",
        "\n",
        "\n",
        "# Create and run a SQL query for the taxi_data from 2014, April, 01\n",
        "query = bq.Query('(SELECT pickup_datetime,  pickup_latitude, pickup_longitude FROM `bigquery-public-data.new_york.tlc_yellow_trips_2014` WHERE pickup_datetime>=\"2014-04-01 00:00:00\" AND pickup_datetime<=\"2014-04-01 23:59:59\" LIMIT 10000)')\n",
        "output_options = bq.QueryOutput.table(use_cache=False)\n",
        "result = query.execute(output_options=output_options).result()\n",
        "\n",
        "# Convert to DataFrame\n",
        "yellow_taxi = result.to_dataframe()\n",
        "yellow_taxi.drop(columns=[\"pickup_datetime\"], inplace=True)\n",
        "yellow_taxi.rename(columns={\"pickup_latitude\":\"lat\", \"pickup_longitude\":\"lon\"}, inplace=True)\n",
        "\n",
        "yellow_taxi.to_csv(\"yellowtaxi_locations_01042014.csv\", index=False)\n",
        "yellow_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.747577</td>\n",
              "      <td>-73.990733</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.734370</td>\n",
              "      <td>-73.990940</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.773287</td>\n",
              "      <td>-73.957660</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.726040</td>\n",
              "      <td>-73.989547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.767507</td>\n",
              "      <td>-73.966557</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         lat        lon\n",
              "0  40.747577 -73.990733\n",
              "1  40.734370 -73.990940\n",
              "2  40.773287 -73.957660\n",
              "3  40.726040 -73.989547\n",
              "4  40.767507 -73.966557"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JJw-u8b4JQ2Y",
        "colab_type": "text"
      },
      "source": [
        "##### Process Raw Uber Taxi Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sRe8sBdSDqBM",
        "colab_type": "code",
        "outputId": "f5a36601-2152-46bf-fc68-8ed744a917db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "# data avaiable here: https://www.kaggle.com/fivethirtyeight/uber-pickups-in-new-york-city#uber-raw-data-apr14.csv\n",
        "uber_taxi = pd.read_csv('uber-raw-data-apr14.csv.zip', parse_dates=['Date/Time'])\n",
        "uber_taxi = uber_taxi[['Date/Time', 'Lat', 'Lon']]\n",
        "uber_taxi = uber_taxi[ (uber_taxi['Date/Time'].dt.month==4) & (uber_taxi['Date/Time'].dt.day==1)]\n",
        "uber_taxi.drop(columns=[\"Date/Time\"], inplace=True)\n",
        "uber_taxi.rename(columns={\"Lat\":\"lat\", \"Lon\":\"lon\"}, inplace=True)\n",
        "uber_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.7690</td>\n",
              "      <td>-73.9549</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.7267</td>\n",
              "      <td>-74.0345</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.7316</td>\n",
              "      <td>-73.9873</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.7588</td>\n",
              "      <td>-73.9776</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.7594</td>\n",
              "      <td>-73.9722</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       lat      lon\n",
              "0  40.7690 -73.9549\n",
              "1  40.7267 -74.0345\n",
              "2  40.7316 -73.9873\n",
              "3  40.7588 -73.9776\n",
              "4  40.7594 -73.9722"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 26
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6AMIjRn4Co-k",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "uber_taxi.to_csv(\"uber_locations_01042014.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hf8V2QPOHanI",
        "colab_type": "text"
      },
      "source": [
        "#### Data Load and Preporcessing\n",
        "\n",
        "* This section reads data for Yellow Taxis and Uber Taxis and performs the following transformations on it:\n",
        "> 1. Attach categorical column **type** to both datasets. **type** can be *yellow_taxi* or *uber_taxi*.\n",
        "> 2. Merge both dataset to create single taxi dataframe called **data**. \n",
        "> 3. Remove *null* and *0* valued latitude and longitudes"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rlBZR9t3D4zc",
        "colab_type": "text"
      },
      "source": [
        "##### Read Yellow Taxi Location Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CV3H3ipXEQU5",
        "colab_type": "code",
        "outputId": "c2a4ad22-5cff-4fed-8867-a1b0fdc6c1ff",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#read processed yellow taxi locations data\n",
        "yellow_taxi = pd.read_csv(\"taxi_location_data_processed/yellowtaxi_locations_01042014.csv\")\n",
        "yellow_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.747577</td>\n",
              "      <td>-73.990733</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.734370</td>\n",
              "      <td>-73.990940</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.773287</td>\n",
              "      <td>-73.957660</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.726040</td>\n",
              "      <td>-73.989547</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.767507</td>\n",
              "      <td>-73.966557</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         lat        lon\n",
              "0  40.747577 -73.990733\n",
              "1  40.734370 -73.990940\n",
              "2  40.773287 -73.957660\n",
              "3  40.726040 -73.989547\n",
              "4  40.767507 -73.966557"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 220
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hqgXg2gfEkXX",
        "colab_type": "code",
        "outputId": "8d60b0a0-281c-4194-a1c1-806ee1207a05",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#find out total Yellow Taxi rides\n",
        "print(\"Total Yellow Taxi Rides: {}\".format(yellow_taxi.shape[0]))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Yellow Taxi Rides: 10000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gL5LW50PGfV0",
        "colab_type": "code",
        "outputId": "3c1e0744-5c7c-4013-c4c4-b4dbdb9af66b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#assing id = 1 for uber cabs\n",
        "yellow_taxi = yellow_taxi.assign(type=\"yellow_taxi\")\n",
        "yellow_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.747577</td>\n",
              "      <td>-73.990733</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.734370</td>\n",
              "      <td>-73.990940</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.773287</td>\n",
              "      <td>-73.957660</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.726040</td>\n",
              "      <td>-73.989547</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.767507</td>\n",
              "      <td>-73.966557</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         lat        lon         type\n",
              "0  40.747577 -73.990733  yellow_taxi\n",
              "1  40.734370 -73.990940  yellow_taxi\n",
              "2  40.773287 -73.957660  yellow_taxi\n",
              "3  40.726040 -73.989547  yellow_taxi\n",
              "4  40.767507 -73.966557  yellow_taxi"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 222
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_s4ZljhRERhw",
        "colab_type": "text"
      },
      "source": [
        "##### Read Uber Location Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "84z2X22WE3RI",
        "colab_type": "code",
        "outputId": "452aea48-8dcd-400c-f26e-980749da70db",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#read processed uber locations data\n",
        "uber_taxi = pd.read_csv(\"taxi_location_data_processed/uber_locations_01042014.csv\")\n",
        "uber_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.7690</td>\n",
              "      <td>-73.9549</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.7267</td>\n",
              "      <td>-74.0345</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.7316</td>\n",
              "      <td>-73.9873</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.7588</td>\n",
              "      <td>-73.9776</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.7594</td>\n",
              "      <td>-73.9722</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       lat      lon\n",
              "0  40.7690 -73.9549\n",
              "1  40.7267 -74.0345\n",
              "2  40.7316 -73.9873\n",
              "3  40.7588 -73.9776\n",
              "4  40.7594 -73.9722"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 223
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "toBJ5_xkE--j",
        "colab_type": "code",
        "outputId": "15c5d959-c5e9-4617-dd96-19c5dca20c1b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#find out total uber rides\n",
        "print(\"Total Uber Rides: {}\".format(uber_taxi.shape[0]))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Uber Rides: 14546\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DZvLQipaFl8D",
        "colab_type": "code",
        "outputId": "4c977e6a-248b-449f-e947-3896bb8742bb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#assing id = 1 for uber cabs\n",
        "uber_taxi = uber_taxi.assign(type=\"uber_taxi\")\n",
        "uber_taxi.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.7690</td>\n",
              "      <td>-73.9549</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.7267</td>\n",
              "      <td>-74.0345</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.7316</td>\n",
              "      <td>-73.9873</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.7588</td>\n",
              "      <td>-73.9776</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.7594</td>\n",
              "      <td>-73.9722</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "       lat      lon       type\n",
              "0  40.7690 -73.9549  uber_taxi\n",
              "1  40.7267 -74.0345  uber_taxi\n",
              "2  40.7316 -73.9873  uber_taxi\n",
              "3  40.7588 -73.9776  uber_taxi\n",
              "4  40.7594 -73.9722  uber_taxi"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 225
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "O6Ku5OKEFeW9",
        "colab_type": "text"
      },
      "source": [
        "##### Create All Taxis"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xsWFdu-nFdQD",
        "colab_type": "code",
        "outputId": "6ecbbd5e-9e3d-45cf-a2c3-83242c164ba4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "#merge both datasets to generate single dataframe \n",
        "data = pd.concat([yellow_taxi, uber_taxi], axis=0)\n",
        "data.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.747577</td>\n",
              "      <td>-73.990733</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.734370</td>\n",
              "      <td>-73.990940</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.773287</td>\n",
              "      <td>-73.957660</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.726040</td>\n",
              "      <td>-73.989547</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.767507</td>\n",
              "      <td>-73.966557</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "         lat        lon         type\n",
              "0  40.747577 -73.990733  yellow_taxi\n",
              "1  40.734370 -73.990940  yellow_taxi\n",
              "2  40.773287 -73.957660  yellow_taxi\n",
              "3  40.726040 -73.989547  yellow_taxi\n",
              "4  40.767507 -73.966557  yellow_taxi"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 226
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6cPrTE7nBhym",
        "colab_type": "code",
        "outputId": "a6111b0c-3b27-4b53-94c5-41f6cf5aa423",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#check total number of rides\n",
        "print(\"Total Rides: {}\".format(data.shape[0]))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Rides: 24546\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "99ShWoAxG-wP",
        "colab_type": "code",
        "outputId": "4785c2a0-4190-4f08-acc6-842f3f351989",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 70
        }
      },
      "source": [
        "#validate if merge took place correctly by checking value counts for ID.\n",
        "data.type.value_counts()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "uber_taxi      14546\n",
              "yellow_taxi    10000\n",
              "Name: type, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 228
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AKah-ox-BXKs",
        "colab_type": "code",
        "outputId": "7f101b3a-e104-459e-eba9-d31d43fa2b70",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "#remove invalid lat, lon (if any)\n",
        "data = data[~(data.lat==0)]\n",
        "data = data[~(data.lon==0)]\n",
        "data = data[~(data.lat.isna())]\n",
        "data = data[~(data.lon.isna())]\n",
        "\n",
        "#check total number of rides\n",
        "print(\"Total Rides after cleanup: {}\".format(data.shape[0]))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Total Rides after cleanup: 24415\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bHcNy4lUPvuC",
        "colab_type": "text"
      },
      "source": [
        "#### Calulating Market Potential for Socially Defined Areas\n",
        "\n",
        "* In this section, we will be calculating the **Market Potential (MP)** for each area defined by zip codes, housing, etc. \n",
        "\n",
        "* This section focuses on finding out if there are any variations in the calculated **MP** when *different* area definitions are used to calculate the same metric.\n",
        "\n",
        "* All the shapefiles used can be found here: [Data Source](https://https://geodata.lib.berkeley.edu/)\n",
        "> Areas by:\n",
        "> 1. [Zip Codes](https://geodata.lib.berkeley.edu/download/file/nyu-2451-34509-geojson.json)\n",
        "> 2. [Housing](https://geodata.lib.berkeley.edu/download/file/nyu-2451-34204-geojson.json)\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oOpKcWTXP3mb",
        "colab_type": "code",
        "outputId": "4c42d421-bec2-4a63-c5b3-e1193ade1dc0",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        }
      },
      "source": [
        "#  setting up CRS for geospatial geometries. Find out more about it here: https://www.w3.org/2015/spatial/wiki/Coordinate_Reference_Systems\n",
        "crs = {'init': 'epsg:4326'}\n",
        "\n",
        "# convert pandas dataframe to geopandas dataframe\n",
        "data_gdf = gpd.GeoDataFrame(data, crs=crs, geometry=[Point(xy) for xy in zip(data.lon, data.lat)])\n",
        "\n",
        "# base_url for different area shapfiles for NYC\n",
        "url=\"https://geodata.lib.berkeley.edu/download/file/\"\n",
        "\n",
        "# select these shapefiles to work with. Each file represents a different area definition\n",
        "area_shape_files={\n",
        "    \n",
        "    \"nyu-2451-34509-geojson.json\":\"zip_code\",\n",
        "    \"nyu-2451-34204-geojson.json\":\"housing\"\n",
        "\n",
        "}\n",
        "\n",
        "# iterate over each area file [ you can extract the code out of the for loop to see how data transforms at each line. ]\n",
        "\n",
        "for i in area_shape_files.keys():\n",
        "  try:\n",
        "    area_gdf = gpd.read_file(url+i, crs=crs) # read geoJSON shapefile for area i\n",
        "    area_gdf = area_gdf[['id','geometry']] # use only columns id, geometry\n",
        "\n",
        "    #__spatially join our based rides data with each area. Finds out which ride happened in which area\n",
        "    data_gdf = sjoin(area_gdf, data_gdf, how='inner', op='contains') \n",
        "\n",
        "    #beautifying area ids\n",
        "    data_gdf[\"id\"] = data_gdf[\"id\"].apply(lambda x: area_shape_files[i]+\"_\"+x[15:])\n",
        "\n",
        "    #some cosmetic changes\n",
        "    data_gdf.drop(columns=[\"index_right\"], inplace=True) \n",
        "    data_gdf.rename(columns={\"id\":area_shape_files[i]+\"_\"+\"id\", \"geometry\":area_shape_files[i]+\"_\"+\"geometry\"}, inplace=True)\n",
        "\n",
        "    \n",
        "    #re-attach geos to compute over again.\n",
        "    data_gdf['geometry']=[Point(xy) for xy in zip(data_gdf.lon, data_gdf.lat)]\n",
        "\n",
        "  except:\n",
        "   print(\"Error in File: {}\".format(i))\n",
        "\n",
        "#cosmetic changes and conversion back to pandas dataframe\n",
        "data_gdf.drop(columns=[\"geometry\"], inplace=True)\n",
        "data = pd.DataFrame(data_gdf)\n",
        "data.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>housing_id</th>\n",
              "      <th>housing_geometry</th>\n",
              "      <th>election_id</th>\n",
              "      <th>election_geometry</th>\n",
              "      <th>zip_code_id</th>\n",
              "      <th>zip_code_geometry</th>\n",
              "      <th>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>zip_code_129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "      <td>40.716000</td>\n",
              "      <td>-73.934200</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>zip_code_129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "      <td>40.716100</td>\n",
              "      <td>-73.934300</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.721200</td>\n",
              "      <td>-73.938300</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.723605</td>\n",
              "      <td>-73.938512</td>\n",
              "      <td>yellow_taxi</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.724500</td>\n",
              "      <td>-73.936400</td>\n",
              "      <td>uber_taxi</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  housing_id  ...         type\n",
              "0  housing_1  ...    uber_taxi\n",
              "0  housing_1  ...    uber_taxi\n",
              "0  housing_1  ...    uber_taxi\n",
              "0  housing_1  ...  yellow_taxi\n",
              "0  housing_1  ...    uber_taxi\n",
              "\n",
              "[5 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 114
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uQQ_6ZXICOzo",
        "colab_type": "code",
        "outputId": "775feb20-38de-4e37-cbab-437234924034",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 364
        }
      },
      "source": [
        "#re-arraning columns for ease of understanding\n",
        "column_order = [\"type\",\"lat\", \"lon\"] + [x for x in data.columns if x not in [\"type\",\"lat\", \"lon\"]]\n",
        "data = data[column_order]\n",
        "data.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>type</th>\n",
              "      <th>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>housing_id</th>\n",
              "      <th>housing_geometry</th>\n",
              "      <th>election_id</th>\n",
              "      <th>election_geometry</th>\n",
              "      <th>zip_code_id</th>\n",
              "      <th>zip_code_geometry</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>40.716000</td>\n",
              "      <td>-73.934200</td>\n",
              "      <td>nyu_2451_34204.1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>nyu_2451_34206.4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>nyu_2451_34509.129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>40.716100</td>\n",
              "      <td>-73.934300</td>\n",
              "      <td>nyu_2451_34204.1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>nyu_2451_34206.4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>nyu_2451_34509.129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>40.721200</td>\n",
              "      <td>-73.938300</td>\n",
              "      <td>nyu_2451_34204.1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>nyu_2451_34206.3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>nyu_2451_34509.140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>40.723605</td>\n",
              "      <td>-73.938512</td>\n",
              "      <td>nyu_2451_34204.1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>nyu_2451_34206.3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>nyu_2451_34509.140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>40.724500</td>\n",
              "      <td>-73.936400</td>\n",
              "      <td>nyu_2451_34204.1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>nyu_2451_34206.3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>nyu_2451_34509.140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "          type  ...                                  zip_code_geometry\n",
              "0    uber_taxi  ...  MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...\n",
              "0    uber_taxi  ...  MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...\n",
              "0    uber_taxi  ...  MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...\n",
              "0  yellow_taxi  ...  MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...\n",
              "0    uber_taxi  ...  MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...\n",
              "\n",
              "[5 rows x 9 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 103
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "esxx2jhdPs-a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#calclating M.P fo each area type\n",
        "housing = computeMPPerPolygon(data, \"housing\")\n",
        "zip_code = computeMPPerPolygon(data, \"zip_code\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "R68jK75vwTmM",
        "colab_type": "text"
      },
      "source": [
        "#### Calulating Market Potential for Hexagons\n",
        "\n",
        "* In this section, we restrict our analysis to a particular **shape** to understand and find out if it helps solve the Shape problem in MAUP. \n",
        "\n",
        "* Here, we are using HEXAGONS as the shape. Also, we analyse Market Potential at different **scales** to find out wether restricting shape helps or not.\n",
        "\n",
        "* We'll be using Uber's H3 library to get the hexagons for each lat, lng. More information on it can be found here: [Uber H3](https://https://github.com/uber/h3)\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Pj7BiI3MsTP-",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# various hexagon resolutions to be used\n",
        "resolutions = [7,8,9]\n",
        "for i in resolutions:\n",
        "  data[\"h3_id_\"+str(i)] = geoToHex(data[['lat', 'lon']].values, i)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BL82EUb-H5iG",
        "colab_type": "code",
        "outputId": "bd01e8d1-373a-48db-94f6-2d6d9d5b1ded",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 467
        }
      },
      "source": [
        "data.head()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>housing_id</th>\n",
              "      <th>housing_geometry</th>\n",
              "      <th>election_id</th>\n",
              "      <th>election_geometry</th>\n",
              "      <th>zip_code_id</th>\n",
              "      <th>zip_code_geometry</th>\n",
              "      <th>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "      <th>h3_id_7</th>\n",
              "      <th>h3_id_8</th>\n",
              "      <th>h3_id_9</th>\n",
              "      <th>h3_id_10</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>zip_code_129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "      <td>40.716000</td>\n",
              "      <td>-73.934200</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>872a100d1ffffff</td>\n",
              "      <td>882a100d13fffff</td>\n",
              "      <td>892a100d137ffff</td>\n",
              "      <td>8a2a100d134ffff</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_4465</td>\n",
              "      <td>MULTIPOLYGON (((-73.92412 40.71769, -73.92410 ...</td>\n",
              "      <td>zip_code_129</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95627 ...</td>\n",
              "      <td>40.716100</td>\n",
              "      <td>-73.934300</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>872a100d1ffffff</td>\n",
              "      <td>882a100d13fffff</td>\n",
              "      <td>892a100d137ffff</td>\n",
              "      <td>8a2a100d134ffff</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.721200</td>\n",
              "      <td>-73.938300</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>872a100d1ffffff</td>\n",
              "      <td>882a100d11fffff</td>\n",
              "      <td>892a100d107ffff</td>\n",
              "      <td>8a2a100d106ffff</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.723605</td>\n",
              "      <td>-73.938512</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>872a100d1ffffff</td>\n",
              "      <td>882a100d11fffff</td>\n",
              "      <td>892a100d117ffff</td>\n",
              "      <td>8a2a100d115ffff</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>housing_1</td>\n",
              "      <td>MULTIPOLYGON (((-73.92390 40.71978, -73.92397 ...</td>\n",
              "      <td>election_3609</td>\n",
              "      <td>MULTIPOLYGON (((-73.92781 40.72506, -73.92920 ...</td>\n",
              "      <td>zip_code_140</td>\n",
              "      <td>MULTIPOLYGON (((-73.95568 40.72394, -73.95615 ...</td>\n",
              "      <td>40.724500</td>\n",
              "      <td>-73.936400</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>872a100d1ffffff</td>\n",
              "      <td>882a100d11fffff</td>\n",
              "      <td>892a100d117ffff</td>\n",
              "      <td>8a2a100d116ffff</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "  housing_id  ...         h3_id_10\n",
              "0  housing_1  ...  8a2a100d134ffff\n",
              "0  housing_1  ...  8a2a100d134ffff\n",
              "0  housing_1  ...  8a2a100d106ffff\n",
              "0  housing_1  ...  8a2a100d115ffff\n",
              "0  housing_1  ...  8a2a100d116ffff\n",
              "\n",
              "[5 rows x 13 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 189
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vnDu7GPu53U0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#calculatin M.Ps per hexagonal area\n",
        "resolutions = [7,8,9]\n",
        "for i in resolutions:\n",
        "  computeMPPerHexagon(data, \"h3_id_\"+str(i))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0i-s1Wa79sgV",
        "colab_type": "text"
      },
      "source": [
        "Using these data files, we plotted the results on our platform. You can use any other visualisation tool too such as Folium, Plotly, etc.\n",
        "\n",
        "![alt text](https://)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "tOJKfEga2B7n",
        "colab_type": "text"
      },
      "source": [
        "#### Infered Areas"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KMDTlOuV2EC8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.cluster import OPTICS, DBSCAN\n",
        "from geopy.distance import great_circle\n",
        "from shapely.geometry import MultiPoint\n",
        "import hdbscan\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UwX52oDpxjcB",
        "colab_type": "code",
        "outputId": "b9793c7d-f383-44f5-be0c-e822a1ceba93",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 230
        }
      },
      "source": [
        "coords = data[['lat', 'lon']].values\n",
        "kms_per_radian = 6371.0088\n",
        "epsilon = 100\n",
        "db = OPTICS(max_eps=epsilon, min_samples = 2, min_cluster_size=5, algorithm='ball_tree', metric='haversine').fit(np.radians(coords))\n",
        "cluster_labels = db.labels_\n",
        "data[\"cluster_id\"] = cluster_labels\n",
        "data.cluster_id.value_counts()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "-1       10520\n",
              " 146        49\n",
              " 110        22\n",
              " 1521       20\n",
              " 145        18\n",
              "         ...  \n",
              " 376         5\n",
              " 360         5\n",
              " 344         5\n",
              " 312         5\n",
              " 0           5\n",
              "Name: cluster_id, Length: 1884, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 196
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "quE_QKGuE1gg",
        "colab_type": "code",
        "outputId": "0f36f43d-7fe6-4efb-f9c1-3d4b8ea6de3f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 520
        }
      },
      "source": [
        "x = data\n",
        "coords = np.radians(x[['lat', 'lon']].values)\n",
        "clusterer = hdbscan.HDBSCAN(metric='haversine', min_cluster_size=15,cluster_selection_epsilon=0.03/6371, min_samples=2)\n",
        "clusterer.fit(coords)\n",
        "cluster_labels = clusterer.labels_\n",
        "x[\"cluster_id\"] = cluster_labels\n",
        "print(\"Noise Total: {} out of {} @ {}%\".format(x[x.cluster_id==-1].shape[0], x.shape[0], x[x.cluster_id==-1].shape[0]*100/x.shape[0]))\n",
        "x = x[x.cluster_id!=-1]\n",
        "x.cluster_id.value_counts().value_counts().plot(kind='barh', figsize=(20, 8))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Noise Total: 4517 out of 24415 @ 18.500921564611918%\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.axes._subplots.AxesSubplot at 0x7fbd225fd390>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 250
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "99x9ygLVLZ_c",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "data[\"cluster_id\"] = cluster_labels"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_u30hIHkKFBS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from shapely.geometry import Polygon, MultiPolygon\n",
        "from scipy.spatial import ConvexHull\n",
        "from shapely.geometry import MultiPoint\n",
        "\n",
        "\n",
        "\n",
        "x = pd.DataFrame(data.groupby(\"cluster_id\").apply(getPolygon)).reset_index()\n",
        "data = pd.merge(data, x, on = \"cluster_id\",how=\"left\")\n",
        "data.dropna(inplace=True)\n",
        "data.rename(columns={0:\"infered_area_geometry\", \"cluster_id\":\"infered_area_id\"}, inplace=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TqcNgeDvLlii",
        "colab_type": "code",
        "outputId": "1a890832-1932-46c3-83b2-527a16620b80",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        }
      },
      "source": [
        "data"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "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>lat</th>\n",
              "      <th>lon</th>\n",
              "      <th>type</th>\n",
              "      <th>infered_area_id</th>\n",
              "      <th>infered_area_geometry</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>40.747577</td>\n",
              "      <td>-73.990733</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>-1</td>\n",
              "      <td>POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>40.734370</td>\n",
              "      <td>-73.990940</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>359</td>\n",
              "      <td>POLYGON ((-73.989254 40.734418, -73.989983 40....</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>40.773287</td>\n",
              "      <td>-73.957660</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>257</td>\n",
              "      <td>POLYGON ((-73.9594 40.7752, -73.9603 40.7738, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>40.726040</td>\n",
              "      <td>-73.989547</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>150</td>\n",
              "      <td>POLYGON ((-73.9892 40.7263, -73.9903 40.7257, ...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>40.767507</td>\n",
              "      <td>-73.966557</td>\n",
              "      <td>yellow_taxi</td>\n",
              "      <td>226</td>\n",
              "      <td>POLYGON ((-73.9661 40.7688, -73.966942 40.7673...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24410</th>\n",
              "      <td>40.721900</td>\n",
              "      <td>-73.992000</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>-1</td>\n",
              "      <td>POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24411</th>\n",
              "      <td>40.726100</td>\n",
              "      <td>-74.002700</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>320</td>\n",
              "      <td>POLYGON ((-74.00291 40.728152, -74.003668 40.7...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24412</th>\n",
              "      <td>40.736400</td>\n",
              "      <td>-73.992600</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>321</td>\n",
              "      <td>POLYGON ((-73.992401 40.738121, -73.99386 40.7...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24413</th>\n",
              "      <td>40.714900</td>\n",
              "      <td>-73.940500</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>-1</td>\n",
              "      <td>POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>24414</th>\n",
              "      <td>40.749300</td>\n",
              "      <td>-73.976800</td>\n",
              "      <td>uber_taxi</td>\n",
              "      <td>-1</td>\n",
              "      <td>POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>24415 rows × 5 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "             lat  ...                              infered_area_geometry\n",
              "0      40.747577  ...  POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...\n",
              "1      40.734370  ...  POLYGON ((-73.989254 40.734418, -73.989983 40....\n",
              "2      40.773287  ...  POLYGON ((-73.9594 40.7752, -73.9603 40.7738, ...\n",
              "3      40.726040  ...  POLYGON ((-73.9892 40.7263, -73.9903 40.7257, ...\n",
              "4      40.767507  ...  POLYGON ((-73.9661 40.7688, -73.966942 40.7673...\n",
              "...          ...  ...                                                ...\n",
              "24410  40.721900  ...  POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...\n",
              "24411  40.726100  ...  POLYGON ((-74.00291 40.728152, -74.003668 40.7...\n",
              "24412  40.736400  ...  POLYGON ((-73.992401 40.738121, -73.99386 40.7...\n",
              "24413  40.714900  ...  POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...\n",
              "24414  40.749300  ...  POLYGON ((-74.1148 40.3082, -73.417 40.7735, -...\n",
              "\n",
              "[24415 rows x 5 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 254
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jKhblOxVLmuv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x = computeMPPerPolygon(data, \"infered_area\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "z4YDTmKbMUcU",
        "colab_type": "text"
      },
      "source": [
        "#### Anaylsis of M.P\n",
        "\n",
        "* This section focuses on analysing the Market Potential based on different areas obtained so far.\n",
        "\n",
        "* We will be finding top 5 Market Potential Areas by the different area definitions we have made so far."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EBGQIuLQMYUM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "zip_code = pd.read_csv(\"results/zip_code.csv\")\n",
        "housing = pd.read_csv(\"results/housing.csv\")\n",
        "h3_id_7 = pd.read_csv(\"results/h3_id_7.csv\")\n",
        "h3_id_8 = pd.read_csv(\"results/h3_id_8.csv\")\n",
        "h3_id_9 = pd.read_csv(\"results/h3_id_9.csv\")\n",
        "infered = pd.read_csv(\"results/infered_area.csv\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hCOxuQVwOxWZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "housing = cutoff(housing, 10)\n",
        "zip_code = cutoff(zip_code, 10)\n",
        "h3_id_7 = cutoff(h3_id_7, 10)\n",
        "h3_id_8 = cutoff(h3_id_8, 10)\n",
        "h3_id_9 = cutoff(h3_id_9, 10)\n",
        "infered = cutoff(infered, 10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zFEjtCxTNZ6N",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "housing.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_housing.csv\", index=False)\n",
        "zip_code.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_zip_code.csv\", index=False)\n",
        "h3_id_7.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_hexagon_7.csv\", index=False)\n",
        "h3_id_8.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_hexagon_8.csv\", index=False)\n",
        "h3_id_9.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_hexagon_9.csv\", index=False)\n",
        "infered.nlargest(5, 'market_potential').to_csv(\"results/top5_mp_areas_by_infered_area.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}