medium-sf.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import geopandas as gpd\n",
    "from shapely.geometry import Point\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "%matplotlib inline\n",
    "import folium\n",
    "import folium.plugins\n",
    "import branca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#https://movement.uber.com/cities/san_francisco/downloads/speeds?lang=en-US&tp[y]=2019&tp[q]=4\n",
    "uber_data= pd.read_csv(r\"movement-speeds-hourly-san-francisco-12-2019.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 26725227 entries, 0 to 26725226\n",
      "Data columns (total 13 columns):\n",
      "year                 int64\n",
      "month                int64\n",
      "day                  int64\n",
      "hour                 int64\n",
      "utc_timestamp        object\n",
      "segment_id           object\n",
      "start_junction_id    object\n",
      "end_junction_id      object\n",
      "osm_way_id           int64\n",
      "osm_start_node_id    int64\n",
      "osm_end_node_id      int64\n",
      "speed_mph_mean       float64\n",
      "speed_mph_stddev     float64\n",
      "dtypes: float64(2), int64(7), object(4)\n",
      "memory usage: 2.6+ GB\n"
     ]
    }
   ],
   "source": [
    "uber_data.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train, test = train_test_split(uber_data, test_size=0.005)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files (x86)\\Microsoft Visual Studio\\Shared\\Anaconda3_64\\lib\\site-packages\\pandas\\core\\frame.py:3694: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  errors=errors)\n"
     ]
    },
    {
     "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>hour</th>\n",
       "      <th>utc_timestamp</th>\n",
       "      <th>osm_way_id</th>\n",
       "      <th>speed_mph_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>4193591</th>\n",
       "      <td>5</td>\n",
       "      <td>2019-12-17T13:00:00.000Z</td>\n",
       "      <td>514417090</td>\n",
       "      <td>60.398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10843245</th>\n",
       "      <td>14</td>\n",
       "      <td>2019-12-04T22:00:00.000Z</td>\n",
       "      <td>288065247</td>\n",
       "      <td>16.057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18706754</th>\n",
       "      <td>19</td>\n",
       "      <td>2019-12-16T03:00:00.000Z</td>\n",
       "      <td>415873749</td>\n",
       "      <td>26.952</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20228321</th>\n",
       "      <td>3</td>\n",
       "      <td>2019-12-29T11:00:00.000Z</td>\n",
       "      <td>397146707</td>\n",
       "      <td>20.114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11531724</th>\n",
       "      <td>11</td>\n",
       "      <td>2019-12-07T19:00:00.000Z</td>\n",
       "      <td>184255674</td>\n",
       "      <td>15.068</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          hour             utc_timestamp  osm_way_id  speed_mph_mean\n",
       "4193591      5  2019-12-17T13:00:00.000Z   514417090          60.398\n",
       "10843245    14  2019-12-04T22:00:00.000Z   288065247          16.057\n",
       "18706754    19  2019-12-16T03:00:00.000Z   415873749          26.952\n",
       "20228321     3  2019-12-29T11:00:00.000Z   397146707          20.114\n",
       "11531724    11  2019-12-07T19:00:00.000Z   184255674          15.068"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.drop(['year', 'month', 'day','segment_id', 'start_junction_id', 'end_junction_id','osm_start_node_id', 'osm_end_node_id', 'speed_mph_stddev'], axis='columns', inplace=True)\n",
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>@id</th>\n",
       "      <th>NHS</th>\n",
       "      <th>bicycle</th>\n",
       "      <th>hgv</th>\n",
       "      <th>hgv:national_network</th>\n",
       "      <th>highway</th>\n",
       "      <th>lanes</th>\n",
       "      <th>maxspeed</th>\n",
       "      <th>name</th>\n",
       "      <th>...</th>\n",
       "      <th>oneway:vehicle</th>\n",
       "      <th>covered</th>\n",
       "      <th>bridge:structure</th>\n",
       "      <th>colour</th>\n",
       "      <th>historic</th>\n",
       "      <th>start_date</th>\n",
       "      <th>crossing</th>\n",
       "      <th>operator</th>\n",
       "      <th>geometry</th>\n",
       "      <th>osm_way_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5071582</td>\n",
       "      <td>5071582</td>\n",
       "      <td>STRAHNET</td>\n",
       "      <td>no</td>\n",
       "      <td>designated</td>\n",
       "      <td>yes</td>\n",
       "      <td>motorway</td>\n",
       "      <td>5</td>\n",
       "      <td>55 mph</td>\n",
       "      <td>James Lick Freeway</td>\n",
       "      <td>...</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>None</td>\n",
       "      <td>LINESTRING (-122.3989008 37.7169496, -122.3989...</td>\n",
       "      <td>582</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>1 rows × 168 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        id      @id       NHS bicycle         hgv hgv:national_network  \\\n",
       "0  5071582  5071582  STRAHNET      no  designated                  yes   \n",
       "\n",
       "    highway lanes maxspeed                name    ...     oneway:vehicle  \\\n",
       "0  motorway     5   55 mph  James Lick Freeway    ...               None   \n",
       "\n",
       "  covered bridge:structure colour historic start_date crossing operator  \\\n",
       "0    None             None   None     None       None     None     None   \n",
       "\n",
       "                                            geometry osm_way_id  \n",
       "0  LINESTRING (-122.3989008 37.7169496, -122.3989...        582  \n",
       "\n",
       "[1 rows x 168 columns]"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#http://overpass-turbo.eu/s/Rq6\n",
    "osm_links = gpd.read_file('sf.geojson')\n",
    "osm_links['osm_way_id']=osm_links['id'].str[4:]\n",
    "osm_links.osm_way_id=osm_links.osm_way_id.astype(int)\n",
    "osm_links.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['id',\n",
       " '@id',\n",
       " 'NHS',\n",
       " 'bicycle',\n",
       " 'hgv',\n",
       " 'hgv:national_network',\n",
       " 'highway',\n",
       " 'lanes',\n",
       " 'maxspeed',\n",
       " 'name',\n",
       " 'oneway',\n",
       " 'ref',\n",
       " 'source:hgv:national_network',\n",
       " 'lanes:backward',\n",
       " 'lanes:forward',\n",
       " 'lcn_ref',\n",
       " 'tiger:cfcc',\n",
       " 'tiger:county',\n",
       " 'tiger:name_base',\n",
       " 'tiger:name_type',\n",
       " 'tiger:reviewed',\n",
       " 'cycleway:right',\n",
       " 'turn:lanes',\n",
       " 'destination',\n",
       " 'destination:ref',\n",
       " 'bridge',\n",
       " 'layer',\n",
       " 'old_ref',\n",
       " 'sidewalk',\n",
       " 'source:imagery',\n",
       " 'surface',\n",
       " 'junction',\n",
       " 'destination:street',\n",
       " 'junction:ref',\n",
       " 'toll',\n",
       " 'maxspeed:advisory',\n",
       " 'history',\n",
       " 'maxheight',\n",
       " 'tiger:name_base_1',\n",
       " 'tiger:name_type_1',\n",
       " 'destination:lanes',\n",
       " 'destination:ref:lanes',\n",
       " 'foot',\n",
       " 'cycleway',\n",
       " 'placement:backward',\n",
       " 'turn:lanes:forward',\n",
       " 'turn:lanes:backward',\n",
       " 'FIXME',\n",
       " 'source',\n",
       " 'parking:lanes:right',\n",
       " 'trolley_wire',\n",
       " 'cycleway:left',\n",
       " 'lanes:both_ways',\n",
       " 'turn:lanes:both_ways',\n",
       " 'hgv:state_network',\n",
       " 'source:hgv:state_network',\n",
       " 'busway',\n",
       " 'lit',\n",
       " 'oneway:bus',\n",
       " 'alt_name',\n",
       " 'name:pronunciation',\n",
       " 'level',\n",
       " 'horse',\n",
       " 'name_1',\n",
       " 'source:maxspeed',\n",
       " 'old_name',\n",
       " 'tiger:name_direction_prefix',\n",
       " 'maxspeed:type',\n",
       " 'pedestrians',\n",
       " 'oneway:psv',\n",
       " 'name:en',\n",
       " 'name:zh',\n",
       " 'cutting',\n",
       " 'name:ja',\n",
       " 'width',\n",
       " 'oneway:bicycle',\n",
       " 'lcn',\n",
       " 'access',\n",
       " 'parking:lane:right',\n",
       " 'name:etymology:wikidata',\n",
       " 'tunnel',\n",
       " 'note:highway',\n",
       " 'note',\n",
       " 'wikidata',\n",
       " 'official_name',\n",
       " 'source:destination',\n",
       " 'parking:lane:both',\n",
       " 'maxweight',\n",
       " 'tiger:name_base_2',\n",
       " 'bicycle:designated',\n",
       " 'oneway:conditional',\n",
       " 'oneway:delivery',\n",
       " 'oneway:taxi',\n",
       " 'destination:name',\n",
       " 'cycleway:both',\n",
       " 'lanes:conditional',\n",
       " 'note:lcn',\n",
       " 'motor_vehicle:conditional',\n",
       " 'placement:forward',\n",
       " 'service',\n",
       " 'maxspeed:hgv',\n",
       " 'maxspeed:trailer',\n",
       " 'bridge:alt_name',\n",
       " 'bridge:name',\n",
       " 'bridge:old_name',\n",
       " 'tiger:zip_right',\n",
       " 'wikipedia',\n",
       " 'tiger:zip_left',\n",
       " 'maxspeed:towing',\n",
       " 'gosm:sig:8CBDE645',\n",
       " 'note:trolley_wire',\n",
       " 'parking:lane:left',\n",
       " 'change',\n",
       " 'not:turn:lanes',\n",
       " 'created_by',\n",
       " 'tiger:name_direction_suffix',\n",
       " 'embankment',\n",
       " 'access:backward',\n",
       " 'building:ref',\n",
       " 'localwiki',\n",
       " 'motor_vehicle',\n",
       " 'bus',\n",
       " 'taxi',\n",
       " 'change:lanes',\n",
       " 'overtaking',\n",
       " 'placement',\n",
       " 'name_base',\n",
       " 'name_type',\n",
       " 'short_name',\n",
       " 'psv:lanes:backward',\n",
       " 'psv:lanes:forward',\n",
       " 'tiger:separated',\n",
       " 'tiger:source',\n",
       " 'tiger:tlid',\n",
       " 'bus:lanes',\n",
       " 'lanes:bus',\n",
       " 'psv',\n",
       " 'access:lanes',\n",
       " 'psv:lanes',\n",
       " 'railway:lanes',\n",
       " 'bus:lanes:backward',\n",
       " 'parking:lanes:both',\n",
       " 'contact:phone',\n",
       " 'diet:halal',\n",
       " 'opening_hours',\n",
       " 'bus:lanes:forward',\n",
       " 'turn',\n",
       " 'height',\n",
       " 'truck',\n",
       " 'vehicle',\n",
       " 'busway:right',\n",
       " 'bicycle:lanes',\n",
       " 'access:lanes:forward',\n",
       " 'busway:forward',\n",
       " 'access:lanes:both_ways',\n",
       " 'parking:lanes:left',\n",
       " 'cycleway:buffer',\n",
       " 'busway:left',\n",
       " 'oneway:vehicle',\n",
       " 'covered',\n",
       " 'bridge:structure',\n",
       " 'colour',\n",
       " 'historic',\n",
       " 'start_date',\n",
       " 'crossing',\n",
       " 'operator',\n",
       " 'geometry',\n",
       " 'osm_way_id']"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "osm_links.columns.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>geometry</th>\n",
       "      <th>osm_way_id</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5071582</td>\n",
       "      <td>LINESTRING (-122.3989008 37.7169496, -122.3989...</td>\n",
       "      <td>582</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>8914982</td>\n",
       "      <td>LINESTRING (-122.3823495 37.7094512, -122.3824...</td>\n",
       "      <td>982</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8914988</td>\n",
       "      <td>LINESTRING (-122.426389 37.7336837, -122.42646...</td>\n",
       "      <td>988</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8914993</td>\n",
       "      <td>LINESTRING (-122.4329133 37.7332794, -122.4327...</td>\n",
       "      <td>993</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8915024</td>\n",
       "      <td>LINESTRING (-122.4073815 37.7361071, -122.4073...</td>\n",
       "      <td>24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                                           geometry  osm_way_id\n",
       "0  5071582  LINESTRING (-122.3989008 37.7169496, -122.3989...         582\n",
       "1  8914982  LINESTRING (-122.3823495 37.7094512, -122.3824...         982\n",
       "2  8914988  LINESTRING (-122.426389 37.7336837, -122.42646...         988\n",
       "3  8914993  LINESTRING (-122.4329133 37.7332794, -122.4327...         993\n",
       "4  8915024  LINESTRING (-122.4073815 37.7361071, -122.4073...          24"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "osm_links.drop([\n",
    " '@id',\n",
    " 'NHS',\n",
    " 'bicycle',\n",
    " 'hgv',\n",
    " 'hgv:national_network',\n",
    " 'highway',\n",
    " 'lanes',\n",
    " 'maxspeed',\n",
    " 'name',\n",
    " 'oneway',\n",
    " 'ref',\n",
    " 'source:hgv:national_network',\n",
    " 'lanes:backward',\n",
    " 'lanes:forward',\n",
    " 'lcn_ref',\n",
    " 'tiger:cfcc',\n",
    " 'tiger:county',\n",
    " 'tiger:name_base',\n",
    " 'tiger:name_type',\n",
    " 'tiger:reviewed',\n",
    " 'cycleway:right',\n",
    " 'turn:lanes',\n",
    " 'destination',\n",
    " 'destination:ref',\n",
    " 'bridge',\n",
    " 'layer',\n",
    " 'old_ref',\n",
    " 'sidewalk',\n",
    " 'source:imagery',\n",
    " 'surface',\n",
    " 'junction',\n",
    " 'destination:street',\n",
    " 'junction:ref',\n",
    " 'toll',\n",
    " 'maxspeed:advisory',\n",
    " 'history',\n",
    " 'maxheight',\n",
    " 'tiger:name_base_1',\n",
    " 'tiger:name_type_1',\n",
    " 'destination:lanes',\n",
    " 'destination:ref:lanes',\n",
    " 'foot',\n",
    " 'cycleway',\n",
    " 'placement:backward',\n",
    " 'turn:lanes:forward',\n",
    " 'turn:lanes:backward',\n",
    " 'FIXME',\n",
    " 'source',\n",
    " 'parking:lanes:right',\n",
    " 'trolley_wire',\n",
    " 'cycleway:left',\n",
    " 'lanes:both_ways',\n",
    " 'turn:lanes:both_ways',\n",
    " 'hgv:state_network',\n",
    " 'source:hgv:state_network',\n",
    " 'busway',\n",
    " 'lit',\n",
    " 'oneway:bus',\n",
    " 'alt_name',\n",
    " 'name:pronunciation',\n",
    " 'level',\n",
    " 'horse',\n",
    " 'name_1',\n",
    " 'source:maxspeed',\n",
    " 'old_name',\n",
    " 'tiger:name_direction_prefix',\n",
    " 'maxspeed:type',\n",
    " 'pedestrians',\n",
    " 'oneway:psv',\n",
    " 'name:en',\n",
    " 'name:zh',\n",
    " 'cutting',\n",
    " 'name:ja',\n",
    " 'width',\n",
    " 'oneway:bicycle',\n",
    " 'lcn',\n",
    " 'access',\n",
    " 'parking:lane:right',\n",
    " 'name:etymology:wikidata',\n",
    " 'tunnel',\n",
    " 'note:highway',\n",
    " 'note',\n",
    " 'wikidata',\n",
    " 'official_name',\n",
    " 'source:destination',\n",
    " 'parking:lane:both',\n",
    " 'maxweight',\n",
    " 'tiger:name_base_2',\n",
    " 'bicycle:designated',\n",
    " 'oneway:conditional',\n",
    " 'oneway:delivery',\n",
    " 'oneway:taxi',\n",
    " 'destination:name',\n",
    " 'cycleway:both',\n",
    " 'lanes:conditional',\n",
    " 'note:lcn',\n",
    " 'motor_vehicle:conditional',\n",
    " 'placement:forward',\n",
    " 'service',\n",
    " 'maxspeed:hgv',\n",
    " 'maxspeed:trailer',\n",
    " 'bridge:alt_name',\n",
    " 'bridge:name',\n",
    " 'bridge:old_name',\n",
    " 'tiger:zip_right',\n",
    " 'wikipedia',\n",
    " 'tiger:zip_left',\n",
    " 'maxspeed:towing',\n",
    " 'gosm:sig:8CBDE645',\n",
    " 'note:trolley_wire',\n",
    " 'parking:lane:left',\n",
    " 'change',\n",
    " 'not:turn:lanes',\n",
    " 'created_by',\n",
    " 'tiger:name_direction_suffix',\n",
    " 'embankment',\n",
    " 'access:backward',\n",
    " 'building:ref',\n",
    " 'localwiki',\n",
    " 'motor_vehicle',\n",
    " 'bus',\n",
    " 'taxi',\n",
    " 'change:lanes',\n",
    " 'overtaking',\n",
    " 'placement',\n",
    " 'name_base',\n",
    " 'name_type',\n",
    " 'short_name',\n",
    " 'psv:lanes:backward',\n",
    " 'psv:lanes:forward',\n",
    " 'tiger:separated',\n",
    " 'tiger:source',\n",
    " 'tiger:tlid',\n",
    " 'bus:lanes',\n",
    " 'lanes:bus',\n",
    " 'psv',\n",
    " 'access:lanes',\n",
    " 'psv:lanes',\n",
    " 'railway:lanes',\n",
    " 'bus:lanes:backward',\n",
    " 'parking:lanes:both',\n",
    " 'contact:phone',\n",
    " 'diet:halal',\n",
    " 'opening_hours',\n",
    " 'bus:lanes:forward',\n",
    " 'turn',\n",
    " 'height',\n",
    " 'truck',\n",
    " 'vehicle',\n",
    " 'busway:right',\n",
    " 'bicycle:lanes',\n",
    " 'access:lanes:forward',\n",
    " 'busway:forward',\n",
    " 'access:lanes:both_ways',\n",
    " 'parking:lanes:left',\n",
    " 'cycleway:buffer',\n",
    " 'busway:left',\n",
    " 'oneway:vehicle',\n",
    " 'covered',\n",
    " 'bridge:structure',\n",
    " 'colour',\n",
    " 'historic',\n",
    " 'start_date',\n",
    " 'crossing',\n",
    " 'operator'], axis='columns', inplace=True)\n",
    "osm_links.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'geopandas.geodataframe.GeoDataFrame'>\n",
      "RangeIndex: 6356 entries, 0 to 6355\n",
      "Data columns (total 3 columns):\n",
      "id            6356 non-null object\n",
      "geometry      6356 non-null object\n",
      "osm_way_id    6356 non-null int32\n",
      "dtypes: int32(1), object(2)\n",
      "memory usage: 124.2+ KB\n"
     ]
    }
   ],
   "source": [
    "osm_links.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 133627 entries, 4193591 to 11429863\n",
      "Data columns (total 4 columns):\n",
      "hour              133627 non-null int64\n",
      "utc_timestamp     133627 non-null object\n",
      "osm_way_id        133627 non-null int64\n",
      "speed_mph_mean    133627 non-null float64\n",
      "dtypes: float64(1), int64(2), object(1)\n",
      "memory usage: 5.1+ MB\n"
     ]
    }
   ],
   "source": [
    "test.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Program Files (x86)\\Microsoft Visual Studio\\Shared\\Anaconda3_64\\lib\\site-packages\\ipykernel_launcher.py:2: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/indexing.html#indexing-view-versus-copy\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "osm_links['osm_way_id'] = osm_links['osm_way_id'].astype(str)\n",
    "test['osm_way_id'] = test['osm_way_id'].astype(str)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "ubersf = pd.merge(osm_links, test, left_on='id', right_on='osm_way_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>id</th>\n",
       "      <th>geometry</th>\n",
       "      <th>osm_way_id_x</th>\n",
       "      <th>hour</th>\n",
       "      <th>utc_timestamp</th>\n",
       "      <th>osm_way_id_y</th>\n",
       "      <th>speed_mph_mean</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>5071582</td>\n",
       "      <td>LINESTRING (-122.3989008 37.7169496, -122.3989...</td>\n",
       "      <td>582</td>\n",
       "      <td>16</td>\n",
       "      <td>2019-12-29T00:00:00.000Z</td>\n",
       "      <td>5071582</td>\n",
       "      <td>61.913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>5071582</td>\n",
       "      <td>LINESTRING (-122.3989008 37.7169496, -122.3989...</td>\n",
       "      <td>582</td>\n",
       "      <td>13</td>\n",
       "      <td>2019-12-05T21:00:00.000Z</td>\n",
       "      <td>5071582</td>\n",
       "      <td>59.601</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>8914988</td>\n",
       "      <td>LINESTRING (-122.426389 37.7336837, -122.42646...</td>\n",
       "      <td>988</td>\n",
       "      <td>19</td>\n",
       "      <td>2019-12-15T03:00:00.000Z</td>\n",
       "      <td>8914988</td>\n",
       "      <td>12.953</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>8914988</td>\n",
       "      <td>LINESTRING (-122.426389 37.7336837, -122.42646...</td>\n",
       "      <td>988</td>\n",
       "      <td>17</td>\n",
       "      <td>2019-12-04T01:00:00.000Z</td>\n",
       "      <td>8914988</td>\n",
       "      <td>20.800</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>8914988</td>\n",
       "      <td>LINESTRING (-122.426389 37.7336837, -122.42646...</td>\n",
       "      <td>988</td>\n",
       "      <td>8</td>\n",
       "      <td>2019-12-11T16:00:00.000Z</td>\n",
       "      <td>8914988</td>\n",
       "      <td>14.887</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        id                                           geometry osm_way_id_x  \\\n",
       "0  5071582  LINESTRING (-122.3989008 37.7169496, -122.3989...          582   \n",
       "1  5071582  LINESTRING (-122.3989008 37.7169496, -122.3989...          582   \n",
       "2  8914988  LINESTRING (-122.426389 37.7336837, -122.42646...          988   \n",
       "3  8914988  LINESTRING (-122.426389 37.7336837, -122.42646...          988   \n",
       "4  8914988  LINESTRING (-122.426389 37.7336837, -122.42646...          988   \n",
       "\n",
       "   hour             utc_timestamp osm_way_id_y  speed_mph_mean  \n",
       "0    16  2019-12-29T00:00:00.000Z      5071582          61.913  \n",
       "1    13  2019-12-05T21:00:00.000Z      5071582          59.601  \n",
       "2    19  2019-12-15T03:00:00.000Z      8914988          12.953  \n",
       "3    17  2019-12-04T01:00:00.000Z      8914988          20.800  \n",
       "4     8  2019-12-11T16:00:00.000Z      8914988          14.887  "
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ubersf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2b34c223518>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "ubersf.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'geopandas.geodataframe.GeoDataFrame'>\n",
      "Int64Index: 25868 entries, 0 to 25867\n",
      "Data columns (total 7 columns):\n",
      "id                25868 non-null object\n",
      "geometry          25868 non-null object\n",
      "osm_way_id_x      25868 non-null object\n",
      "hour              25868 non-null int64\n",
      "utc_timestamp     25868 non-null object\n",
      "osm_way_id_y      25868 non-null object\n",
      "speed_mph_mean    25868 non-null float64\n",
      "dtypes: float64(1), int64(1), object(5)\n",
      "memory usage: 1.6+ MB\n"
     ]
    }
   ],
   "source": [
    "ubersf.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "ubersf.drop(['id', 'osm_way_id_x', 'hour','osm_way_id_y'], axis='columns', inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "ubersf.to_file(\"uber_sf.geojson\", encoding='utf-8', driver=\"GeoJSON\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Ready for kepler.gl"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}