San Francisco Business Registrations.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "This Notebook creates the landing image seen in the slides for JupyterCon\n",
    "\"\"\"\n",
    "\n",
    "\n",
    "import geopandas as gpd\n",
    "import ipywidgets\n",
    "import pandas as pd\n",
    "import pydeck as pdk\n",
    "\n",
    "## Load in our data sets\n",
    "df = pd.read_csv('https://raw.githubusercontent.com/ajduberstein/sf_growth/master/public/data/business.csv')\n",
    "\n",
    "# Handle null values which may not serialize into deck.gl\n",
    "df = df.fillna(0)\n",
    "\n",
    "# Load in neighborhood boundaries for San Francisco\n",
    "gdf = gpd.read_file('https://raw.githubusercontent.com/ajduberstein/sf_growth/master/public/data/neighborhoods.geojson')\n",
    "\n",
    "# Zoom to the middle 80% of the business data (presumably centered in San Francisco)\n",
    "viewport = pdk.data_utils.compute_view(df[['lng', 'lat']], 0.8)\n",
    "\n",
    "\n",
    "# Create two scatter plots, one which will turn orange when new businesses are added\n",
    "# and one which cumulatively tracks all businesses opened until that point\n",
    "\n",
    "PURPLE = [154, 18, 179, 200]\n",
    "ORANGE = [255, 98, 0]\n",
    "\n",
    "def make_scatterplot(data, fill_color):\n",
    "    return pdk.Layer(\n",
    "        \"ScatterplotLayer\",\n",
    "        data=data,\n",
    "        pickable=True,\n",
    "        get_position=['lng', 'lat'],\n",
    "        radius_min_pixels=1,\n",
    "        # Radius of the point in meters\n",
    "        get_radius=10,\n",
    "        radius_width_min_pixels=2,\n",
    "        get_fill_color=fill_color,\n",
    "    )\n",
    "\n",
    "base_layer = make_scatterplot(df, PURPLE)\n",
    "new_layer = make_scatterplot([], ORANGE)\n",
    "\n",
    "# Define the GeoJsonLayer for the neighborhood boundaries\n",
    "\n",
    "neighborhood_boundaries = pdk.Layer(\n",
    "    \"GeoJsonLayer\",\n",
    "    data=gdf,\n",
    "    stroked=True,\n",
    "    filled=False,\n",
    "    line_width_min_pixels=2,\n",
    "    get_radius=1\n",
    ")\n",
    "\n",
    "# Merge the visualization components\n",
    "layers = [new_layer, base_layer, neighborhood_boundaries]\n",
    "deck = pdk.Deck(\n",
    "    layers,\n",
    "    initial_view_state=viewport,\n",
    "    map_provider=None,\n",
    "    tooltip={'text': '{business_name}, opened {start_date}'}\n",
    ")\n",
    "\n",
    "# Render\n",
    "display(deck.show())\n",
    "year = ipywidgets.HTML()\n",
    "display(year)\n",
    "\n",
    "# Play the data forward in time, from 1950-2018 (takes 68 seconds)\n",
    "def play():\n",
    "    global deck\n",
    "    global df\n",
    "    global year\n",
    "    for i in range(int(df.start_date.min()), int(df.start_date.max()) + 1):\n",
    "        deck.layers[0].data = df[df['start_date'] == i]\n",
    "        deck.layers[1].data = df[df['start_date'] < i]\n",
    "        year.value = f'<b>{i}</b>'\n",
    "        deck.update()\n",
    "        time.sleep(1)\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "play()"
   ]
  }
 ],
 "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": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}