FinalProject_group060.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# COGS 108 - Final Project"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Overview"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "For this project, we are using two datasets which include the street light point locations as well as the crime happened in San Diego. We convert the addresses from the crime dataset with recent 180 days to the light location with latitude and longitude which from the street light dataset. We separated the data by zip code with the first three digits. And convert all watts to lumens with the\n",
    "average brightness (lumens) per square meter of zip code. By the end of the process, we will examine the relationships between the number of streetlights in San Diego and the number of crimes that happened.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Names"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "        * Adrian Werner\n",
    "        * Qilong Li\n",
    "        * Zijian wang\n",
    "        * MengYuan Shi\n",
    "        * Zhirong Lin"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Research Question"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We are interested in finding out how the number of streetlights in San Diego affects the number of crimes on that street. Is it possible to estimate how the number of streetlights in San Diego that affects the crimes on the streets? In this project, we aim to examine the relationships between the crimes and the streetlights by using datasets. \n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Background and Prior Work"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "There are various factors that affect the crime rate, and in the United States today it is thought that crime’s negative effects are unfairly felt by disadvantaged populations. As criminality and socioeconomics become linked, governmental aid or infrastructure has the potential to redistribute some of this burden. The streetlight here serves as a model of said infrastructure, in that it is a manifestation of public funds to create greater visibility and potentially safety. \n",
    "\n",
    "A 2019 Crime Lab study focused on the city of New York and the relationship between providing new streetlights to low-income housing developments and the crime rate. Their research question came from the idea that reducing crime was, in the past, thought to be an extremely complicated matter that required extensive public resources and funding. They proposed that perhaps, instead, a slight environmental modification could have some effect. Adding streetlights did show a positive correlation with a reduced crime rate, according to their study. \n",
    "\n",
    "While some critics of the proposal suggest that creating more or rather more powerful streetlights in an attempt to reduce the crime rate is an unreliable use of a valuable resource (electricity), America’s largest cities are proving the opposite. The so-called “smart streetlight” is a multi-tool light, camera, and alarm system with significantly greater luminance than the traditional streetlight. It has also helped the city of Los Angeles bring down their energy costs nearly $9 million. The LAPD reported that in areas with converted “smart streetlights,” the crime rate has dropped. Local news in San Diego, already in early January 2020, has reported this technology being introduced there.\n",
    " \n",
    "- References (include links):\n",
    "\n",
    "    1)Aaron Chalfin, Benjamin Hansen, Jason Lerner, and Lucie Parker, “Reducing Crime Through Environmental Design: Evidence from a Randomized Experiment of Street Lighting in New York City,” Crime Lab New York, April 24, 2019, https://doi.org/10.1257/rct.4186-1.0.\n",
    "\n",
    "    2)Teena Maddox, “How LA is now saving $9M a year with LED streetlights and converting them into EV charging stations,” TechRepublic, July 7, 2016, https://www.techrepublic.com/article/how-la-is-now-saving-9m-a-year-with-led-streetlights-and-converting-them-into-ev-charging-stations/.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Hypothesis"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1.We predict that the lack of streetlights on the street may increase the number of crimes that happen on that street.\n",
    "\n",
    "2.We also predict that the greater distribution of streetlights may also increase the number of violent crimes (i.e. homicide).\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Dataset(s)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Datasets 1: \n",
    "\n",
    "- Dataset Name: \n",
    "    - Streetlight point location in San Diego\n",
    "\n",
    "- Link to the dataset:\n",
    "    - https://data.sandiego.gov/datasets/streetlight_inventory/ \n",
    "\n",
    "- Number of observations: \n",
    "    - 61230 data\n",
    "\n",
    "For this dataset, it contains the type of the street light, the model for the street light, as well as the wattage and voltage of the street light, moreover, we have the description of each street light location which including the latitude and longitude. As the area is so big, we try to go up to the top three areas which have the most street lights This dataset includes coordinates for each light.\n",
    "\n",
    "- For our research question, we chose to use:\n",
    "    1. street_wattage;\n",
    "    2. Light location which latitude and longitude\n",
    "\n",
    "- Problems with this data: \n",
    "    1. The regions of the areas are too big.\n",
    "    2. The observation for this dataset is too much, it’s hard to get the part of data we want to use for our project.\n",
    "\n",
    "\n",
    "\n",
    "Datasets 2:\n",
    "\n",
    "- Dataset Name: \n",
    "    - Crimes happen in San Diego\n",
    "\n",
    "- Link to the dataset: \n",
    "    - https://www.sandag.org/index.asp?classid=14&subclassid=21&projectid=446&fuseaction=projects.detail\n",
    "\n",
    "- Number of observations: \n",
    "    - 61688 data\n",
    "\n",
    "For this Dataset it includes the variables: the date for the crime happening in a certain area, the address of the blocks, zip code for that area and as well as the community. This dataset is the crimes that have happened in the most recent 180 days in San Diego regional. We try to separate data by zip code which is up to first 3 digits.\n",
    "\n",
    "- For our research question, we chose to use:\n",
    "    1. Zip code;\n",
    "    2. Region for only San Diego;\n",
    "    3. Addresses\n",
    "\n",
    "- Problems with this dataset: \n",
    "    1. There is no correct observations we have because It’s a csv dataset and the table will be updated weekly. \n",
    "    2. The data does not only contain the San Diego region.\n",
    "    3. Data only contains the recent 180 days and doesn't have the data for a year or more to compare with.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 134,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: geocoder in ./.local/lib/python3.6/site-packages (1.38.1)\n",
      "Requirement already satisfied: six in /opt/conda/lib/python3.6/site-packages (from geocoder) (1.12.0)\n",
      "Requirement already satisfied: ratelim in ./.local/lib/python3.6/site-packages (from geocoder) (0.1.6)\n",
      "Requirement already satisfied: future in ./.local/lib/python3.6/site-packages (from geocoder) (0.18.2)\n",
      "Requirement already satisfied: requests in /opt/conda/lib/python3.6/site-packages (from geocoder) (2.12.4)\n",
      "Requirement already satisfied: click in ./.local/lib/python3.6/site-packages (from geocoder) (7.1.1)\n",
      "Requirement already satisfied: decorator in /opt/conda/lib/python3.6/site-packages (from ratelim->geocoder) (4.3.0)\n",
      "Requirement already satisfied: geopy in ./.local/lib/python3.6/site-packages (1.21.0)\n",
      "Requirement already satisfied: geographiclib<2,>=1.49 in ./.local/lib/python3.6/site-packages (from geopy) (1.50)\n",
      "Requirement already satisfied: openpyxl in ./.local/lib/python3.6/site-packages (3.0.3)\n",
      "Requirement already satisfied: jdcal in ./.local/lib/python3.6/site-packages (from openpyxl) (1.4.1)\n",
      "Requirement already satisfied: et-xmlfile in ./.local/lib/python3.6/site-packages (from openpyxl) (1.0.1)\n",
      "Requirement already satisfied: gmaps in ./.local/lib/python3.6/site-packages (0.9.0)\n",
      "Requirement already satisfied: traitlets>=4.3.0 in /opt/conda/lib/python3.6/site-packages (from gmaps) (4.3.2)\n",
      "Requirement already satisfied: geojson>=2.0.0 in ./.local/lib/python3.6/site-packages (from gmaps) (2.5.0)\n",
      "Requirement already satisfied: six in /opt/conda/lib/python3.6/site-packages (from gmaps) (1.12.0)\n",
      "Requirement already satisfied: ipywidgets>=7.0.0 in /opt/conda/lib/python3.6/site-packages (from gmaps) (7.2.1)\n",
      "Requirement already satisfied: ipython>=5.3.0 in /opt/conda/lib/python3.6/site-packages (from gmaps) (7.2.0)\n",
      "Requirement already satisfied: ipython_genutils in /opt/conda/lib/python3.6/site-packages (from traitlets>=4.3.0->gmaps) (0.2.0)\n",
      "Requirement already satisfied: decorator in /opt/conda/lib/python3.6/site-packages (from traitlets>=4.3.0->gmaps) (4.3.0)\n",
      "Requirement already satisfied: widgetsnbextension~=3.2.0 in /opt/conda/lib/python3.6/site-packages (from ipywidgets>=7.0.0->gmaps) (3.2.1)\n",
      "Requirement already satisfied: nbformat>=4.2.0 in /opt/conda/lib/python3.6/site-packages (from ipywidgets>=7.0.0->gmaps) (4.4.0)\n",
      "Requirement already satisfied: ipykernel>=4.5.1 in /opt/conda/lib/python3.6/site-packages (from ipywidgets>=7.0.0->gmaps) (5.1.0)\n",
      "Requirement already satisfied: setuptools>=18.5 in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (40.6.3)\n",
      "Requirement already satisfied: jedi>=0.10 in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (0.13.2)\n",
      "Requirement already satisfied: pickleshare in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (0.7.5)\n",
      "Requirement already satisfied: prompt_toolkit<2.1.0,>=2.0.0 in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (2.0.7)\n",
      "Requirement already satisfied: pygments in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (2.3.1)\n",
      "Requirement already satisfied: backcall in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (0.1.0)\n",
      "Requirement already satisfied: pexpect in /opt/conda/lib/python3.6/site-packages (from ipython>=5.3.0->gmaps) (4.6.0)\n",
      "Requirement already satisfied: notebook>=4.4.1 in /opt/conda/lib/python3.6/site-packages (from widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (5.7.2)\n",
      "Requirement already satisfied: jsonschema!=2.5.0,>=2.4 in /opt/conda/lib/python3.6/site-packages (from nbformat>=4.2.0->ipywidgets>=7.0.0->gmaps) (3.0.0a3)\n",
      "Requirement already satisfied: jupyter_core in /opt/conda/lib/python3.6/site-packages (from nbformat>=4.2.0->ipywidgets>=7.0.0->gmaps) (4.4.0)\n",
      "Requirement already satisfied: jupyter_client in /opt/conda/lib/python3.6/site-packages (from ipykernel>=4.5.1->ipywidgets>=7.0.0->gmaps) (5.2.4)\n",
      "Requirement already satisfied: tornado>=4.2 in /opt/conda/lib/python3.6/site-packages (from ipykernel>=4.5.1->ipywidgets>=7.0.0->gmaps) (5.1.1)\n",
      "Requirement already satisfied: parso>=0.3.0 in /opt/conda/lib/python3.6/site-packages (from jedi>=0.10->ipython>=5.3.0->gmaps) (0.3.1)\n",
      "Requirement already satisfied: wcwidth in /opt/conda/lib/python3.6/site-packages (from prompt_toolkit<2.1.0,>=2.0.0->ipython>=5.3.0->gmaps) (0.1.7)\n",
      "Requirement already satisfied: ptyprocess>=0.5 in /opt/conda/lib/python3.6/site-packages (from pexpect->ipython>=5.3.0->gmaps) (0.6.0)\n",
      "Requirement already satisfied: Send2Trash in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (1.5.0)\n",
      "Requirement already satisfied: jinja2 in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (2.10)\n",
      "Requirement already satisfied: nbconvert in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (5.6.1)\n",
      "Requirement already satisfied: pyzmq>=17 in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (17.1.2)\n",
      "Requirement already satisfied: terminado>=0.8.1 in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.8.1)\n",
      "Requirement already satisfied: prometheus-client in /opt/conda/lib/python3.6/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.5.0)\n",
      "Requirement already satisfied: attrs>=17.4.0 in /opt/conda/lib/python3.6/site-packages (from jsonschema!=2.5.0,>=2.4->nbformat>=4.2.0->ipywidgets>=7.0.0->gmaps) (18.2.0)\n",
      "Requirement already satisfied: pyrsistent>=0.14.0 in /opt/conda/lib/python3.6/site-packages (from jsonschema!=2.5.0,>=2.4->nbformat>=4.2.0->ipywidgets>=7.0.0->gmaps) (0.14.8)\n",
      "Requirement already satisfied: python-dateutil>=2.1 in /opt/conda/lib/python3.6/site-packages (from jupyter_client->ipykernel>=4.5.1->ipywidgets>=7.0.0->gmaps) (2.7.5)\n",
      "Requirement already satisfied: MarkupSafe>=0.23 in /opt/conda/lib/python3.6/site-packages (from jinja2->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (1.1.0)\n",
      "Requirement already satisfied: entrypoints>=0.2.2 in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.2.3)\n",
      "Requirement already satisfied: testpath in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.4.2)\n",
      "Requirement already satisfied: pandocfilters>=1.4.1 in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (1.4.2)\n",
      "Requirement already satisfied: mistune<2,>=0.8.1 in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.8.4)\n",
      "Requirement already satisfied: defusedxml in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.6.0)\n",
      "Requirement already satisfied: bleach in /opt/conda/lib/python3.6/site-packages (from nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (3.0.2)\n",
      "Requirement already satisfied: webencodings in /opt/conda/lib/python3.6/site-packages (from bleach->nbconvert->notebook>=4.4.1->widgetsnbextension~=3.2.0->ipywidgets>=7.0.0->gmaps) (0.5.1)\n",
      "Requirement already satisfied: sklearn in ./.local/lib/python3.6/site-packages (0.0)\n",
      "Requirement already satisfied: scikit-learn in /opt/conda/lib/python3.6/site-packages (from sklearn) (0.21.3)\n",
      "Requirement already satisfied: scipy>=0.17.0 in /opt/conda/lib/python3.6/site-packages (from scikit-learn->sklearn) (1.1.0)\n",
      "Requirement already satisfied: numpy>=1.11.0 in /opt/conda/lib/python3.6/site-packages (from scikit-learn->sklearn) (1.17.3)\n",
      "Requirement already satisfied: joblib>=0.11 in /opt/conda/lib/python3.6/site-packages (from scikit-learn->sklearn) (0.14.0)\n",
      "Enabling notebook extension jupyter-gmaps/extension...\n",
      "      - Validating: \u001b[32mOK\u001b[0m\n",
      "Enabling notebook extension jupyter-js-widgets/extension...\n",
      "      - Validating: \u001b[32mOK\u001b[0m\r\n"
     ]
    }
   ],
   "source": [
    "! pip install --user geocoder\n",
    "! pip install --user geopy\n",
    "! pip install --user openpyxl\n",
    "! pip install --user gmaps\n",
    "!pip install --user sklearn\n",
    "\n",
    "! jupyter nbextension enable --py  gmaps\n",
    "! jupyter nbextension enable --py  widgetsnbextension"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Imports used in project:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 135,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import sklearn\n",
    "from sklearn.linear_model import LinearRegression\n",
    "\n",
    "import seaborn as sns\n",
    "sns.set()\n",
    "sns.set_context('talk')\n",
    "\n",
    "import geocoder\n",
    "import geopy\n",
    "import gmaps.datasets\n",
    "\n",
    "from geopy.geocoders import Nominatim\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')\n",
    "\n",
    "import patsy\n",
    "import statsmodels.api as sm\n",
    "import scipy.stats as stats\n",
    "from scipy.stats import ttest_ind, chisquare, normaltest"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Our first dataset shows the location of streetlights throughout San Diego:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 136,
   "metadata": {},
   "outputs": [],
   "source": [
    "sl_df = pd.read_csv('http://seshat.datasd.org/streetlight_inventory/streetlight_locations_datasd_v1.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(61229, 8)\n"
     ]
    }
   ],
   "source": [
    "sl_shape = sl_df.shape\n",
    "print(sl_shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 138,
   "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>sap_id</th>\n",
       "      <th>streetlight_model</th>\n",
       "      <th>streetlight_type</th>\n",
       "      <th>streetlight_wattage</th>\n",
       "      <th>streetlight_voltage</th>\n",
       "      <th>lng</th>\n",
       "      <th>lat</th>\n",
       "      <th>location_description</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>SS-015619-001</td>\n",
       "      <td>COBRA</td>\n",
       "      <td>IND</td>\n",
       "      <td>85.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>-117.252424</td>\n",
       "      <td>32.864282</td>\n",
       "      <td>LA JOLLA SHORES DR @ RUETTE MONTE CARLO N/E</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>SS-015617-002</td>\n",
       "      <td>COBRA</td>\n",
       "      <td>IND</td>\n",
       "      <td>85.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>-117.250892</td>\n",
       "      <td>32.872267</td>\n",
       "      <td>LA JOLLA SHORES LN N/O LA JOLLA SHORES DR @ CDS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SS-005044-001</td>\n",
       "      <td>COBRA</td>\n",
       "      <td>IND</td>\n",
       "      <td>85.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>-117.246841</td>\n",
       "      <td>32.881493</td>\n",
       "      <td>BROOKMEAD LN E/O BLACK GOLD RD @ CDS</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>SS-004379-003</td>\n",
       "      <td>COBRA</td>\n",
       "      <td>IND</td>\n",
       "      <td>85.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>-117.248476</td>\n",
       "      <td>32.885452</td>\n",
       "      <td>BLACK GOLD RD E/O LA JOLLA FARMS RD 200' N/S</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SS-004379-004</td>\n",
       "      <td>COBRA</td>\n",
       "      <td>IND</td>\n",
       "      <td>85.0</td>\n",
       "      <td>120.0</td>\n",
       "      <td>-117.247382</td>\n",
       "      <td>32.885273</td>\n",
       "      <td>BLACK GOLD RD E/O LA JOLLA FARMS RD 680' S/S</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "          sap_id streetlight_model streetlight_type  streetlight_wattage  \\\n",
       "0  SS-015619-001             COBRA              IND                 85.0   \n",
       "1  SS-015617-002             COBRA              IND                 85.0   \n",
       "2  SS-005044-001             COBRA              IND                 85.0   \n",
       "3  SS-004379-003             COBRA              IND                 85.0   \n",
       "4  SS-004379-004             COBRA              IND                 85.0   \n",
       "\n",
       "   streetlight_voltage         lng        lat  \\\n",
       "0                120.0 -117.252424  32.864282   \n",
       "1                120.0 -117.250892  32.872267   \n",
       "2                120.0 -117.246841  32.881493   \n",
       "3                120.0 -117.248476  32.885452   \n",
       "4                120.0 -117.247382  32.885273   \n",
       "\n",
       "                              location_description  \n",
       "0      LA JOLLA SHORES DR @ RUETTE MONTE CARLO N/E  \n",
       "1  LA JOLLA SHORES LN N/O LA JOLLA SHORES DR @ CDS  \n",
       "2             BROOKMEAD LN E/O BLACK GOLD RD @ CDS  \n",
       "3     BLACK GOLD RD E/O LA JOLLA FARMS RD 200' N/S  \n",
       "4     BLACK GOLD RD E/O LA JOLLA FARMS RD 680' S/S  "
      ]
     },
     "execution_count": 138,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here is a description of the data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 139,
   "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>field</th>\n",
       "      <th>description</th>\n",
       "      <th>values</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>sap_id</td>\n",
       "      <td>Streetlight ID in SAP</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>streetlight_model</td>\n",
       "      <td>Streetlight model</td>\n",
       "      <td>Ballpark lighting, Barn light, Basketball cour...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>streetlight_type</td>\n",
       "      <td>Streetlight type</td>\n",
       "      <td>HPS, Incandescent, IND, LED, LPS, Mercury vapo...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>streetlight_wattage</td>\n",
       "      <td>Streetlight wattage</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>streetlight_voltage</td>\n",
       "      <td>Streetlight voltage</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>longitude</td>\n",
       "      <td>Longitude</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>latitude</td>\n",
       "      <td>Latitude</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>location_description</td>\n",
       "      <td>Text description of location</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  field                   description  \\\n",
       "0                sap_id         Streetlight ID in SAP   \n",
       "1     streetlight_model             Streetlight model   \n",
       "2      streetlight_type              Streetlight type   \n",
       "3   streetlight_wattage           Streetlight wattage   \n",
       "4   streetlight_voltage           Streetlight voltage   \n",
       "5             longitude                     Longitude   \n",
       "6              latitude                      Latitude   \n",
       "7  location_description  Text description of location   \n",
       "\n",
       "                                              values  \n",
       "0                                                NaN  \n",
       "1  Ballpark lighting, Barn light, Basketball cour...  \n",
       "2  HPS, Incandescent, IND, LED, LPS, Mercury vapo...  \n",
       "3                                                NaN  \n",
       "4                                                NaN  \n",
       "5                                                NaN  \n",
       "6                                                NaN  \n",
       "7                                                NaN  "
      ]
     },
     "execution_count": 139,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_dict = pd.read_csv('http://seshat.datasd.org/streetlight_inventory/streetlight_locations_dictionary_datasd.csv')\n",
    "sl_dict = sl_dict[[\"field\", \"description\", \"values\"]]\n",
    "sl_dict"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the all the different values for the ```streetlight_model``` column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 140,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "COBRA                     39788\n",
       "POSTTOP                    7727\n",
       "SHOEBOX                    3236\n",
       "BALLPARK LIGHTIING         2433\n",
       "GATEWAY                    2259\n",
       "GASLAMP                    1242\n",
       "PENDANT                     830\n",
       "COBRA DOUBLE HEAD           688\n",
       "TEARDROP                    557\n",
       "FLoodLight                  548\n",
       "TENNIS COURT LIGHTING       481\n",
       "BASKETBALL COURT            199\n",
       "BOLLARD                     177\n",
       "POSTTOP-SCRIPPS RANCH       173\n",
       "PATHWAY                     172\n",
       "PARKING LOT                 109\n",
       "SOFFIT                      105\n",
       "MISCELLANEOUS-1HEAD          88\n",
       "GRASS LIGHTING               68\n",
       "POOL LIGHTING                42\n",
       "WALPACK                      36\n",
       "MISCELLANEOUS-MANYHEAD       23\n",
       "DECORATIVE NONSTANDARD       18\n",
       "NO LIGHT HERE                16\n",
       "Chinese Lantern              11\n",
       "SKATEPARK LIGHTING            3\n",
       "BARN LIGHT                    3\n",
       "Name: streetlight_model, dtype: int64"
      ]
     },
     "execution_count": 140,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df['streetlight_model'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "28"
      ]
     },
     "execution_count": 141,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(sl_df['streetlight_model'].unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And here are the all the different values for the ```streetlight_type``` column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 142,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "IND              35139\n",
       "LPS               4749\n",
       "LED               4141\n",
       "HPS               3625\n",
       "UNKNOWN            385\n",
       "OTHER               65\n",
       "INCANDESCENT        11\n",
       "MERCURY VAPOR        9\n",
       "Other                7\n",
       "Unknown              5\n",
       "METAL HALIDE         2\n",
       "Name: streetlight_type, dtype: int64"
      ]
     },
     "execution_count": 142,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df['streetlight_type'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 143,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 143,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(sl_df['streetlight_type'].unique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We will look at the descriptive statistics for the numeric columns, excluding the coordinate data (```lat``` and ```lng```) since they are spatial/geographic data and measures of their central tendency do not necessarily apply."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 144,
   "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>streetlight_wattage</th>\n",
       "      <th>streetlight_voltage</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>50468.000000</td>\n",
       "      <td>39554.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>94.847270</td>\n",
       "      <td>163.595186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>52.702749</td>\n",
       "      <td>106.739893</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>85.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>85.000000</td>\n",
       "      <td>120.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>85.000000</td>\n",
       "      <td>240.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1000.000000</td>\n",
       "      <td>5000.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       streetlight_wattage  streetlight_voltage\n",
       "count         50468.000000         39554.000000\n",
       "mean             94.847270           163.595186\n",
       "std              52.702749           106.739893\n",
       "min               0.000000             0.000000\n",
       "25%              85.000000           120.000000\n",
       "50%              85.000000           120.000000\n",
       "75%              85.000000           240.000000\n",
       "max            1000.000000          5000.000000"
      ]
     },
     "execution_count": 144,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df[['streetlight_wattage', 'streetlight_voltage']].describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We want to know the number of missing data points in the data set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "72326"
      ]
     },
     "execution_count": 145,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df.isna().sum().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And which columns they are found in:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "sap_id                   True\n",
       "streetlight_model        True\n",
       "streetlight_type         True\n",
       "streetlight_wattage      True\n",
       "streetlight_voltage      True\n",
       "lng                     False\n",
       "lat                     False\n",
       "location_description     True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 146,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df.isna().any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "And finally the percentage of missing data points in each column:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "sap_id                  0.221382\n",
       "streetlight_model       0.003217\n",
       "streetlight_type        0.213804\n",
       "streetlight_wattage     0.175750\n",
       "streetlight_voltage     0.353999\n",
       "lng                     0.000000\n",
       "lat                     0.000000\n",
       "location_description    0.213085\n",
       "dtype: float64"
      ]
     },
     "execution_count": 147,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df.isna().mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on this, we can see that all columns, besides ```lat``` and ```lng```, have missing data. This means that coordinate locations for all the streetlights exist, but some are missing the type of streetlight (```streetlight_type```) or the light's wattage (```streetlight_wattage```), for example. \n",
    "\n",
    "However, there is one value in the ```streetlight_model``` column that reads \"NO LIGHT HERE.\" These can be dropped:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 148,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "COBRA                     39788\n",
       "POSTTOP                    7727\n",
       "SHOEBOX                    3236\n",
       "BALLPARK LIGHTIING         2433\n",
       "GATEWAY                    2259\n",
       "GASLAMP                    1242\n",
       "PENDANT                     830\n",
       "COBRA DOUBLE HEAD           688\n",
       "TEARDROP                    557\n",
       "FLoodLight                  548\n",
       "TENNIS COURT LIGHTING       481\n",
       "BASKETBALL COURT            199\n",
       "BOLLARD                     177\n",
       "POSTTOP-SCRIPPS RANCH       173\n",
       "PATHWAY                     172\n",
       "PARKING LOT                 109\n",
       "SOFFIT                      105\n",
       "MISCELLANEOUS-1HEAD          88\n",
       "GRASS LIGHTING               68\n",
       "POOL LIGHTING                42\n",
       "WALPACK                      36\n",
       "MISCELLANEOUS-MANYHEAD       23\n",
       "DECORATIVE NONSTANDARD       18\n",
       "Chinese Lantern              11\n",
       "SKATEPARK LIGHTING            3\n",
       "BARN LIGHT                    3\n",
       "Name: streetlight_model, dtype: int64"
      ]
     },
     "execution_count": 148,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df.drop(sl_df[sl_df['streetlight_model'] == 'NO LIGHT HERE'].index , inplace=True)\n",
    "sl_df['streetlight_model'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "We can consider the different wattages among streetlights, and if conversion is necessary because of the different types (LED vs. incandescent). However, because of the lack of a unit, we do not know if the wattage is given consistently as pure watts, lumens per watts, or uses some conversion factor to simulate the number of watts for LED lighting. Thus, streetlight brightness cannot accurately be considered in our analysis."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "85.0     25918\n",
       "165.0     8428\n",
       "0.0       4919\n",
       "81.0      3672\n",
       "55.0      2864\n",
       "Name: streetlight_wattage, dtype: int64"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df['streetlight_wattage'].value_counts().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 150,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "85.0     25903\n",
       "165.0     8423\n",
       "55.0       368\n",
       "250.0      105\n",
       "150.0       78\n",
       "Name: streetlight_wattage, dtype: int64"
      ]
     },
     "execution_count": 150,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df['streetlight_wattage'][sl_df['streetlight_type'] == 'IND'].value_counts().head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 151,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "81.0     3672\n",
       "250.0      81\n",
       "0.0        65\n",
       "150.0      21\n",
       "55.0       18\n",
       "Name: streetlight_wattage, dtype: int64"
      ]
     },
     "execution_count": 151,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sl_df['streetlight_wattage'][sl_df['streetlight_type'] == 'LED'].value_counts().head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below is the data of San Diego regional crime for the past 180 days, updated weekly on Wednesdays:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 152,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_crime = pd.read_csv('http://www.sandag.org/programs/public_safety/arjis/CrimeData/crimedata.zip', dtype = {'ZipCode':str})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 153,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(61533, 7)\n"
     ]
    }
   ],
   "source": [
    "crime_shape = df_crime.shape\n",
    "print(crime_shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "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>CM_LEGEND</th>\n",
       "      <th>agency</th>\n",
       "      <th>Charge_Description_Orig</th>\n",
       "      <th>activityDate</th>\n",
       "      <th>BLOCK_ADDRESS</th>\n",
       "      <th>ZipCode</th>\n",
       "      <th>community</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>THEFT/LARCENY</td>\n",
       "      <td>OCEANSIDE</td>\n",
       "      <td>PETTY THEFT(All Other Larceny) (M)</td>\n",
       "      <td>12/29/2019 18:11:00</td>\n",
       "      <td>1800  BLOCK COLLEGE BOULEVARD</td>\n",
       "      <td>92056</td>\n",
       "      <td>OCEANSIDE</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>THEFT/LARCENY</td>\n",
       "      <td>CHULA VISTA</td>\n",
       "      <td>PETTY THEFT(Shoplift) (M)</td>\n",
       "      <td>1/2/2020 20:15:00</td>\n",
       "      <td>600  BLOCK PALOMAR STREET</td>\n",
       "      <td>91911</td>\n",
       "      <td>CHULA VISTA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>DUI</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>DUI ALC/0.08 PERCENT (M)</td>\n",
       "      <td>1/1/2020 2:36:00</td>\n",
       "      <td>4000  BLOCK RUEDA DRIVE</td>\n",
       "      <td>92124</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>MOTOR VEHICLE THEFT</td>\n",
       "      <td>CHULA VISTA</td>\n",
       "      <td>TAKE VEHICLE W/O OWNER'S CONSENT/VEHICLE THEFT...</td>\n",
       "      <td>1/6/2020 17:00:00</td>\n",
       "      <td>1200  BLOCK ATWATER STREET</td>\n",
       "      <td>91913</td>\n",
       "      <td>CHULA VISTA</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>DRUGS/ALCOHOL VIOLATIONS</td>\n",
       "      <td>ESCONDIDO</td>\n",
       "      <td>POSSESS CONTROLLED SUBSTANCE (M)</td>\n",
       "      <td>12/28/2019 20:00:00</td>\n",
       "      <td>500 W  BLOCK WASHINGTON AVENUE</td>\n",
       "      <td>92025</td>\n",
       "      <td>ESCONDIDO</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                  CM_LEGEND       agency  \\\n",
       "0             THEFT/LARCENY    OCEANSIDE   \n",
       "1             THEFT/LARCENY  CHULA VISTA   \n",
       "2                       DUI    SAN DIEGO   \n",
       "3       MOTOR VEHICLE THEFT  CHULA VISTA   \n",
       "4  DRUGS/ALCOHOL VIOLATIONS    ESCONDIDO   \n",
       "\n",
       "                             Charge_Description_Orig         activityDate  \\\n",
       "0                 PETTY THEFT(All Other Larceny) (M)  12/29/2019 18:11:00   \n",
       "1                          PETTY THEFT(Shoplift) (M)    1/2/2020 20:15:00   \n",
       "2                           DUI ALC/0.08 PERCENT (M)     1/1/2020 2:36:00   \n",
       "3  TAKE VEHICLE W/O OWNER'S CONSENT/VEHICLE THEFT...    1/6/2020 17:00:00   \n",
       "4                   POSSESS CONTROLLED SUBSTANCE (M)  12/28/2019 20:00:00   \n",
       "\n",
       "                    BLOCK_ADDRESS ZipCode    community  \n",
       "0   1800  BLOCK COLLEGE BOULEVARD   92056    OCEANSIDE  \n",
       "1       600  BLOCK PALOMAR STREET   91911  CHULA VISTA  \n",
       "2         4000  BLOCK RUEDA DRIVE   92124    SAN DIEGO  \n",
       "3      1200  BLOCK ATWATER STREET   91913  CHULA VISTA  \n",
       "4  500 W  BLOCK WASHINGTON AVENUE   92025    ESCONDIDO  "
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 155,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DRUNK IN PUBLIC: ALCOHOL, DRUGS, COMBO OR TOLUENE (M)    4367\n",
       "TAKE VEHICLE W/O OWNER'S CONSENT/VEHICLE THEFT (F)       4253\n",
       "POSSESS CONTROLLED SUBSTANCE (M)                         3809\n",
       "BURGLARY (VEHICLE) (F)                                   3624\n",
       "VANDALISM ($400 OR MORE) (F)                             2605\n",
       "                                                         ... \n",
       "POSS FIREARM W/ALTERED IDENT (F)                            1\n",
       "DISCHARGE FIREARM FROM VEHICLE (F)                          1\n",
       "SEND/POSS OBSCENE MATTER (M)                                1\n",
       "HAVE FIREARM AT PUBLIC/ETC SCHOOL                           1\n",
       "FAIL REG GAIN/LV RESID (F)                                  1\n",
       "Name: Charge_Description_Orig, Length: 572, dtype: int64"
      ]
     },
     "execution_count": 155,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime['Charge_Description_Orig'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Since there are 575 different values in the ```Charge_Description_Orig``` column, this might not be this best way to categorize the crimes. ```CM_LEGEND``` also groups by type of crime, but more broadly:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 156,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DRUGS/ALCOHOL VIOLATIONS    16005\n",
       "THEFT/LARCENY                9370\n",
       "ASSAULT                      7342\n",
       "VEHICLE BREAK-IN/THEFT       6745\n",
       "MOTOR VEHICLE THEFT          4858\n",
       "VANDALISM                    4550\n",
       "BURGLARY                     3425\n",
       "FRAUD                        2935\n",
       "DUI                          2546\n",
       "ROBBERY                      1288\n",
       "SEX CRIMES                   1271\n",
       "WEAPONS                      1019\n",
       "ARSON                         159\n",
       "HOMICIDE                       20\n",
       "Name: CM_LEGEND, dtype: int64"
      ]
     },
     "execution_count": 156,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime['CM_LEGEND'].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 157,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2291"
      ]
     },
     "execution_count": 157,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime.isna().sum().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "CM_LEGEND                  False\n",
       "agency                     False\n",
       "Charge_Description_Orig    False\n",
       "activityDate               False\n",
       "BLOCK_ADDRESS              False\n",
       "ZipCode                     True\n",
       "community                   True\n",
       "dtype: bool"
      ]
     },
     "execution_count": 158,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime.isna().any()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While there are missing data points, they are only for the ```ZipCode``` and ```community``` columns. This means that there is sufficient information for each row. At the least we have the type of crime(```CM_LEGEND```) and location(```BLOCK_ADDRESS```)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 159,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "92101    5238\n",
       "92109    2482\n",
       "92054    2051\n",
       "92020    1611\n",
       "92113    1591\n",
       "         ... \n",
       "92002       1\n",
       "91401       1\n",
       "92072       1\n",
       "92179       1\n",
       "90038       1\n",
       "Name: ZipCode, Length: 149, dtype: int64"
      ]
     },
     "execution_count": 159,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_crime['ZipCode'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here's a plot of our map."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 160,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "028dba289691466bb308678e9e965ebe",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Figure(layout=FigureLayout(height='420px'))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "\n",
    "import gmaps.datasets\n",
    "gmaps.configure(api_key='AIzaSyCWUoHIgK3KNUk8za3Q1iULpMAf-HyionQ')\n",
    "fig = gmaps.figure()\n",
    "\n",
    "# generate some (latitude, longitude) pairs\n",
    "locations = list(zip(sl_df[\"lat\"],sl_df[\"lng\"]))\n",
    "\n",
    "heatmap_layer = gmaps.heatmap_layer(locations)\n",
    "fig.add_layer(heatmap_layer)\n",
    "fig\n",
    "#heat maps for traffic lights we need to pick some hotspots and cold spot to see whether they have the same rates of crime.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## More Cleaning of the crime Data Frame\n",
    "\n",
    "Here is the format to convert addresses to coordinate data. It will have to be incorporated into a function that takes values from ```BLOCK_ADDRESS``` and ```ZipCode``` as input and outputs the latitude and longitude. ```BLOCK_ADDRESS``` and ```ZipCode``` should be comma-separated. (The word 'BLOCK' may have to be removed from the data if it does not return coordinates.) We have performed some more data cleaning since geocoder requires a combination of address and zipcode together wehave combined them into a new column  \"address\". Removed crime that happens during the daytime. We have also removed the word \"BLOCK\" from the address. However, the function that we created in geocoder does not support large data sets due to limited quota per person per day."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 161,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "32.802338146353044,-117.19590289054813\n",
      "(32.802338146353044, -117.19590289054813)\n"
     ]
    }
   ],
   "source": [
    "geolocator = Nominatim(user_agent=\"specify_your_app_name_here\")\n",
    "#location = geolocator.geocode(\"1800 BLOCK COLLEGE BOULEVARD, 92056\")\n",
    "location = geolocator.geocode(\"3410 waco st, 92117\")\n",
    "\n",
    "def geocoder_conv(string): #Note: geocoder causes timeout when running large datasets this function does not work for large sets\n",
    "    geolocator = Nominatim(user_agent=\"specify_your_app_name_here\")\n",
    "    if(geolocator.geocode(string)==None):\n",
    "        return\n",
    "    else:\n",
    "        location= geolocator.geocode(string)\n",
    "        string= str(location.latitude)+ \",\"+str(location.longitude)\n",
    "        return string\n",
    "    \n",
    "string1=geocoder_conv(\"3410 waco st, 92117\")\n",
    "print(string1)\n",
    "print((location.latitude, location.longitude))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Clean the data with keywords block in address to make our geocoder more efficient.\n",
    "we also need to clean the data one more time to leave out the daytime and only account for crimes around time between 6pm- 6am."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_crime = df_crime.assign(address = df_crime['BLOCK_ADDRESS']+\", \" +df_crime['ZipCode'].astype(str))\n",
    "timeout_address = []\n",
    "df_crime.head()\n",
    "\n",
    "#remove block from addresses\n",
    "def standardize_address(string):\n",
    "    \n",
    "    string = string.lower()\n",
    "    string = string.strip()\n",
    "    \n",
    "    if 'block' in string:\n",
    "        word_list = string.split();\n",
    "        output = ' '.join([i for i in word_list if i not in 'block'])\n",
    "        return output\n",
    "    else:\n",
    "        return string\n",
    "\n",
    "    \n",
    "def remove_and_address():\n",
    "    df_test = df_crime[~df_crime.address.str.contains('&')]\n",
    "    return df_test\n",
    "\n",
    "\n",
    "   \n",
    "\n",
    "#\n",
    "df_crime['address'] = df_crime['address'].apply(standardize_address)\n",
    "df_crime = remove_and_address()\n",
    "df_crime['time'] = [x.split(' ')[1] for x in df_crime['activityDate']]\n",
    "df_crime['time'] =  pd.to_datetime(df_crime['time'], format='%H:%M:%S')\n",
    "\n",
    "\n",
    "#from_ts = pd.date.time('06:00:00')                             #filiter time from 0600-1800\n",
    "#to_ts = pd.date.time('06:00:00')\n",
    "#df= df_crime['time'].between_time('06:30', '18:30')\n",
    "#df_crime = df_crime[(df_crime['new_time'] < from_ts) | (df_crime['new_time'] > to_ts)]\n",
    "\n",
    "#df_crime['address']= df_crime['address'].apply(geocoder_conv)\n",
    "start_date = pd.to_datetime('1/1/1900 6:00')\n",
    "end_date = pd.to_datetime('1/1/1900 18:00')\n",
    "df_crime_night=df_crime.loc[(df_crime['time'] < start_date) | (df_crime['time'] > end_date)]\n",
    "#df_crime_day= df_crime.loc[(df_crime['time'] > start_date)]\n",
    "df_crime_day=df_crime.loc[(df_crime['time'] > start_date) & (df_crime['time'] < end_date)]\n",
    "df_crime= df_crime_night\n",
    "df_crime=df_crime.drop(columns=['time'])\n",
    "\n",
    "df_crime = df_crime.reset_index(drop=True)\n",
    "df_crime = df_crime[df_crime['community'].notna()]\n",
    "indexNames = df_crime[ df_crime['CM_LEGEND'] == 'DUI' ].index\n",
    "indexNames1 = df_crime[ df_crime['CM_LEGEND'] == 'FRAUD' ].index\n",
    "indexNames2 = df_crime[ df_crime['Charge_Description_Orig'] == 'SHOPLIFTING' ].index\n",
    "\n",
    "# Delete these row indexes from dataFrame\n",
    "df_crime.drop(indexNames , inplace=True)\n",
    "df_crime.drop(indexNames1 , inplace=True)\n",
    "df_crime.drop(indexNames2 , inplace=True)\n",
    "df_crime\n",
    "#df_crime['community'].value_counts()==1\n",
    "df_crime.to_csv(\"output3.csv\")                         #export to excel file\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Next Step   \n",
    "We export the dataframe to googlesheets and used the addon geocode from googlesheet to convert addresses to coordinates. with two added column of longitude and latitude. we imported the .csv file back to the data. we decided to only use the data points in San Diego, and also remove other crimes that is not revelant to streetlights such as shoplifting, domestic violance ets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 163,
   "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>CM_LEGEND</th>\n",
       "      <th>agency</th>\n",
       "      <th>Charge_Description_Orig</th>\n",
       "      <th>activityDate</th>\n",
       "      <th>BLOCK_ADDRESS</th>\n",
       "      <th>ZipCode</th>\n",
       "      <th>community</th>\n",
       "      <th>address</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>Latitude</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>DRUGS/ALCOHOL VIOLATIONS</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>DRUNK IN PUBLIC: ALCOHOL, DRUGS, COMBO OR TOLU...</td>\n",
       "      <td>10/18/2019 22:10:00</td>\n",
       "      <td>700  BLOCK GRAND AVENUE</td>\n",
       "      <td>92109</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>700 grand avenue, 92109</td>\n",
       "      <td>-117.256047</td>\n",
       "      <td>32.794497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>VEHICLE BREAK-IN/THEFT</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>BURGLARY (VEHICLE) (F)</td>\n",
       "      <td>12/30/2019 4:50:00</td>\n",
       "      <td>8800  BLOCK BETELGEUSE WAY</td>\n",
       "      <td>92126</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>8800 betelgeuse way, 92126</td>\n",
       "      <td>-117.133896</td>\n",
       "      <td>32.933881</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>SIMPLE BATTERY (M)</td>\n",
       "      <td>12/30/2019 1:00:00</td>\n",
       "      <td>200  BLOCK 55TH STREET</td>\n",
       "      <td>92114</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>200 55th street, 92114</td>\n",
       "      <td>-117.076435</td>\n",
       "      <td>32.706388</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>DRUGS/ALCOHOL VIOLATIONS</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>TRANS/ETC CONTROLLED SUBS (F)</td>\n",
       "      <td>1/16/2020 22:00:00</td>\n",
       "      <td>1600  BLOCK ISLAND AVENUE</td>\n",
       "      <td>92101</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>1600 island avenue, 92101</td>\n",
       "      <td>-117.149435</td>\n",
       "      <td>32.710464</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>SEX CRIMES</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>DISORDERLY CONDUCT: PROSTITUTION - SEX FOR COM...</td>\n",
       "      <td>1/30/2020 21:45:00</td>\n",
       "      <td>400  BLOCK HOTEL CIRCLE SOUTH</td>\n",
       "      <td>92108</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>400 hotel circle south, 92108</td>\n",
       "      <td>-117.177008</td>\n",
       "      <td>32.759391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9840</th>\n",
       "      <td>VEHICLE BREAK-IN/THEFT</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>PETTY THEFT(Mot Veh Parts) (M)</td>\n",
       "      <td>10/10/2019 0:25:00</td>\n",
       "      <td>4400  BLOCK WINONA AVENUE</td>\n",
       "      <td>92115</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>4400 winona avenue, 92115</td>\n",
       "      <td>-117.087922</td>\n",
       "      <td>32.756669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9841</th>\n",
       "      <td>VEHICLE BREAK-IN/THEFT</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>BURGLARY (VEHICLE) (F)</td>\n",
       "      <td>10/10/2019 18:15:00</td>\n",
       "      <td>4100  BLOCK GESNER STREET</td>\n",
       "      <td>92117</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>4100 gesner street, 92117</td>\n",
       "      <td>-117.205325</td>\n",
       "      <td>32.791904</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9842</th>\n",
       "      <td>THEFT/LARCENY</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>GRAND THEFT:MONEY/LABOR/PROPERTY (F)</td>\n",
       "      <td>10/10/2019 22:00:00</td>\n",
       "      <td>10600  BLOCK CAMINITO DERECHO</td>\n",
       "      <td>92126</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>10600 caminito derecho, 92126</td>\n",
       "      <td>-117.127503</td>\n",
       "      <td>32.910657</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9843</th>\n",
       "      <td>MOTOR VEHICLE THEFT</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>TAKE VEHICLE W/O OWNER'S CONSENT/VEHICLE THEFT...</td>\n",
       "      <td>10/16/2019 0:01:00</td>\n",
       "      <td>2900  BLOCK LARKIN PLACE</td>\n",
       "      <td>92123</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>2900 larkin place, 92123</td>\n",
       "      <td>-117.134236</td>\n",
       "      <td>32.795814</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9844</th>\n",
       "      <td>VANDALISM</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>VANDALISM ($400 OR MORE) (F)</td>\n",
       "      <td>9/20/2019 1:30:00</td>\n",
       "      <td>1000  BLOCK SAPPHIRE STREET</td>\n",
       "      <td>92109</td>\n",
       "      <td>SAN DIEGO</td>\n",
       "      <td>1000 sapphire street, 92109</td>\n",
       "      <td>-117.255239</td>\n",
       "      <td>32.808579</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>9845 rows × 10 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     CM_LEGEND     agency  \\\n",
       "0     DRUGS/ALCOHOL VIOLATIONS  SAN DIEGO   \n",
       "1       VEHICLE BREAK-IN/THEFT  SAN DIEGO   \n",
       "2                      ASSAULT  SAN DIEGO   \n",
       "3     DRUGS/ALCOHOL VIOLATIONS  SAN DIEGO   \n",
       "4                   SEX CRIMES  SAN DIEGO   \n",
       "...                        ...        ...   \n",
       "9840    VEHICLE BREAK-IN/THEFT  SAN DIEGO   \n",
       "9841    VEHICLE BREAK-IN/THEFT  SAN DIEGO   \n",
       "9842             THEFT/LARCENY  SAN DIEGO   \n",
       "9843       MOTOR VEHICLE THEFT  SAN DIEGO   \n",
       "9844                 VANDALISM  SAN DIEGO   \n",
       "\n",
       "                                Charge_Description_Orig         activityDate  \\\n",
       "0     DRUNK IN PUBLIC: ALCOHOL, DRUGS, COMBO OR TOLU...  10/18/2019 22:10:00   \n",
       "1                                BURGLARY (VEHICLE) (F)   12/30/2019 4:50:00   \n",
       "2                                    SIMPLE BATTERY (M)   12/30/2019 1:00:00   \n",
       "3                         TRANS/ETC CONTROLLED SUBS (F)   1/16/2020 22:00:00   \n",
       "4     DISORDERLY CONDUCT: PROSTITUTION - SEX FOR COM...   1/30/2020 21:45:00   \n",
       "...                                                 ...                  ...   \n",
       "9840                     PETTY THEFT(Mot Veh Parts) (M)   10/10/2019 0:25:00   \n",
       "9841                             BURGLARY (VEHICLE) (F)  10/10/2019 18:15:00   \n",
       "9842               GRAND THEFT:MONEY/LABOR/PROPERTY (F)  10/10/2019 22:00:00   \n",
       "9843  TAKE VEHICLE W/O OWNER'S CONSENT/VEHICLE THEFT...   10/16/2019 0:01:00   \n",
       "9844                       VANDALISM ($400 OR MORE) (F)    9/20/2019 1:30:00   \n",
       "\n",
       "                      BLOCK_ADDRESS  ZipCode  community  \\\n",
       "0           700  BLOCK GRAND AVENUE    92109  SAN DIEGO   \n",
       "1        8800  BLOCK BETELGEUSE WAY    92126  SAN DIEGO   \n",
       "2            200  BLOCK 55TH STREET    92114  SAN DIEGO   \n",
       "3         1600  BLOCK ISLAND AVENUE    92101  SAN DIEGO   \n",
       "4     400  BLOCK HOTEL CIRCLE SOUTH    92108  SAN DIEGO   \n",
       "...                             ...      ...        ...   \n",
       "9840      4400  BLOCK WINONA AVENUE    92115  SAN DIEGO   \n",
       "9841      4100  BLOCK GESNER STREET    92117  SAN DIEGO   \n",
       "9842  10600  BLOCK CAMINITO DERECHO    92126  SAN DIEGO   \n",
       "9843       2900  BLOCK LARKIN PLACE    92123  SAN DIEGO   \n",
       "9844    1000  BLOCK SAPPHIRE STREET    92109  SAN DIEGO   \n",
       "\n",
       "                            address   Longitude   Latitude  \n",
       "0           700 grand avenue, 92109 -117.256047  32.794497  \n",
       "1        8800 betelgeuse way, 92126 -117.133896  32.933881  \n",
       "2            200 55th street, 92114 -117.076435  32.706388  \n",
       "3         1600 island avenue, 92101 -117.149435  32.710464  \n",
       "4     400 hotel circle south, 92108 -117.177008  32.759391  \n",
       "...                             ...         ...        ...  \n",
       "9840      4400 winona avenue, 92115 -117.087922  32.756669  \n",
       "9841      4100 gesner street, 92117 -117.205325  32.791904  \n",
       "9842  10600 caminito derecho, 92126 -117.127503  32.910657  \n",
       "9843       2900 larkin place, 92123 -117.134236  32.795814  \n",
       "9844    1000 sapphire street, 92109 -117.255239  32.808579  \n",
       "\n",
       "[9845 rows x 10 columns]"
      ]
     },
     "execution_count": 163,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\n",
    "df_crime_sd=pd.read_csv('crime_SD.csv')       #get the file back from googlesheets, with newly added columns\n",
    "\n",
    "df_crime_sd=df_crime_sd.drop(df_crime_sd.columns[0], axis=1)\n",
    "\n",
    "\n",
    "\n",
    "df_crime_sd\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Import our data points on gmaps to get the sense of crime distributions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 164,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "4966f211d1ec49d7ac7e9dec2c857525",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Figure(layout=FigureLayout(height='420px'))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "gmaps.configure(api_key='AIzaSyCWUoHIgK3KNUk8za3Q1iULpMAf-HyionQ')\n",
    "crime_map = gmaps.figure()\n",
    "\n",
    "# generate some (latitude, longitude) pairs\n",
    "locations = list(zip(df_crime_sd[\"Latitude\"],df_crime_sd[\"Longitude\"]))\n",
    "\n",
    "heatmap_layer = gmaps.heatmap_layer(locations)\n",
    "crime_map.add_layer(heatmap_layer)\n",
    "crime_map"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "below are the functions to divide our data into sub region. since all the sub region have the same area we can somehow have a \n",
    "sense how  number of crime and streetlights are related to each other."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 165,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "\n",
    "def long_div_list(numdiv):\n",
    "    list1=[]\n",
    "    for i in np.linspace(df_crime_sd['Longitude'].min(), df_crime_sd['Longitude'].max(), num=numdiv):\n",
    "        list1.append(i)\n",
    "    return list1\n",
    "    \n",
    "latitude=[]    \n",
    "def lat_div_list(numdiv):\n",
    "    list2=[]\n",
    "    for i in np.linspace(df_crime_sd['Latitude'].min(), df_crime_sd['Latitude'].max(), num=numdiv):\n",
    "        list2.append(i)\n",
    "    return list2\n",
    "\n",
    "def list_crime_division(numdiv):\n",
    "    long_list= long_div_list(numdiv)\n",
    "    lat_list=lat_div_list(numdiv)\n",
    "    crime_list1=[]\n",
    "    for i in range(len(long_list)-1):\n",
    "        for j in range(len(lat_list)-1):\n",
    "            df=df_crime_sd.loc[(df_crime_sd['Latitude']>lat_list[j])&(df_crime_sd['Latitude']< lat_list[j+1])&(df_crime_sd['Longitude']>long_list[i]) &(df_crime_sd['Longitude']<long_list[i+1])]\n",
    "            crime_list1.append(len(df.index))\n",
    "    return crime_list1\n",
    "\n",
    "def list_light_division(numdiv):\n",
    "    long_list= long_div_list(numdiv)\n",
    "    lat_list=lat_div_list(numdiv)\n",
    "    light_list1=[]\n",
    "    for i in range(len(long_list)-1):\n",
    "        for j in range(len(lat_list)-1):\n",
    "            df=sl_df.loc[(sl_df['lat']>lat_list[j])&(sl_df['lat']< lat_list[j+1])&(sl_df['lng']>long_list[i]) &(sl_df['lng']<long_list[i+1])]\n",
    "            light_list1.append(len(df.index))\n",
    "    return light_list1\n",
    "    \n",
    "    \n",
    "\n",
    "\n",
    "       \n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "we first try to divide with few regions to see if there are any relations. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 166,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            OLS Regression Results                            \n",
      "==============================================================================\n",
      "Dep. Variable:                  light   R-squared:                       0.924\n",
      "Model:                            OLS   Adj. R-squared:                  0.917\n",
      "Method:                 Least Squares   F-statistic:                     145.4\n",
      "Date:                Thu, 19 Mar 2020   Prob (F-statistic):           4.59e-08\n",
      "Time:                        22:45:04   Log-Likelihood:                -124.30\n",
      "No. Observations:                  14   AIC:                             252.6\n",
      "Df Residuals:                      12   BIC:                             253.9\n",
      "Df Model:                           1                                         \n",
      "Covariance Type:            nonrobust                                         \n",
      "==============================================================================\n",
      "                 coef    std err          t      P>|t|      [0.025      0.975]\n",
      "------------------------------------------------------------------------------\n",
      "Intercept   1250.7619    562.816      2.222      0.046      24.492    2477.032\n",
      "crime          4.3841      0.364     12.056      0.000       3.592       5.176\n",
      "==============================================================================\n",
      "Omnibus:                        1.225   Durbin-Watson:                   1.568\n",
      "Prob(Omnibus):                  0.542   Jarque-Bera (JB):                0.231\n",
      "Skew:                           0.296   Prob(JB):                        0.891\n",
      "Kurtosis:                       3.211   Cond. No.                     1.74e+03\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 1.74e+03. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "\n",
    "crime2=list_crime_division(5)\n",
    "light2= list_light_division(5)\n",
    "crime_vs_light2=pd.DataFrame(list(zip(crime2,light2)),columns=['crime','light'])\n",
    "crime_vs_light2= crime_vs_light2.loc[(crime_vs_light2['crime']!=0) | (crime_vs_light2['light']!=0)]\n",
    "#crime_vs_light2\n",
    "#pd.plotting.scatter_matrix(crime_vs_light2) \n",
    "\n",
    "\n",
    "\n",
    "\n",
    "func,axis = plt.subplots()\n",
    "\n",
    "plt.ylabel('crime')\n",
    "plt.xlabel('light')\n",
    "\n",
    "x=np.array(light2)\n",
    "y=np.array(crime2)\n",
    "\n",
    "linreg= LinearRegression()\n",
    "x=x.reshape(-1,1)\n",
    "linreg.fit(x,y)\n",
    "y_pred=linreg.predict(x)\n",
    "plt.scatter(x,y)\n",
    "plt.plot(x,y_pred,color='red')\n",
    "\n",
    "outcome_1, predictors_1 = patsy.dmatrices('light ~ crime', crime_vs_light2)\n",
    "mod_1 = sm.OLS(outcome_1, predictors_1)\n",
    "res_1 = mod_1.fit()\n",
    "print(res_1.summary())\n",
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "looks like there is some positive relations between number of lights and crime. we are now dividing the sub region \n",
    "into 100 blocks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 167,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f2cb76db320>]"
      ]
     },
     "execution_count": 167,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAb4AAAEaCAYAAABuADIRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nO3deVxV1f4//tc+TAIyiYApCihC4gCVA6hJqWiDQzZoda+I2nSvU6aWt65m3X7VRw1SrK+ZmY03zVRCS9NMKsAhzQkHQBSnyzwcJpnO/v1BHN3uc+AcBM6wX8/Hw4ey9trrrHUO+GbtNQmiKIogIiJSCJWpK0BERNSeGPiIiEhRGPiIiEhRGPiIiEhRGPiIiEhRGPiIiEhRTBb4UlNTsXjxYowdOxahoaEYMWIEZs+ejXPnzsnyJicnY/LkyRgwYAAiIiKwdOlSqNVqWb6Kigq89dZbGD58OAYMGIBHH30UP//8s87XN7RMIiKyLoKp1vHNnTsXJSUlePDBB9GrVy8UFBRg/fr1yMjIwBdffIGwsDAAwMGDBzFjxgyMGjUKTz75JPLy8rBy5Ur4+vri66+/hkp1I3ZPnz4dp0+fxsKFC+Hr64tt27YhMTERa9euRWRkpDafMWUaSxRFtOQdFYTG+1v80haLbVde25XaboBtB+RtFwRAaLzYHvUwVeArLCyEp6enJE2tVmPUqFEIDw9HfHw8AODxxx9HXV0dtm7dqg1IycnJmDFjBuLi4vDQQw8BAJKSkvDcc89hzZo1iIqKAtAQhJ5++mmUlJTgxx9/1L6OoWW2hEYjorCw3Oj73NwcAQClpVUtfm1LxbYrr+1KbTfAtgPytnt6doRK1X6Bz2SPOm8NegDg6uoKPz8/5OTkAAByc3Nx8uRJTJw4UdILGzZsGHx8fLB7925t2p49e+Di4oJRo0Zp0wRBwKRJk5CVlYXMzEyjyyQiIutjVpNbioqKkJGRgd69ewMA0tPTAUD79c2CgoKQkZGh/TojIwOBgYGyx5TBwcGSsowpk4iIrI+tqSvQSBRFLFmyBBqNBjNnzgQAlJSUAADc3Nxk+d3c3HD69Gnt1yUlJfD399eZ7+ayjCmzJQThRnfeGLa2Nn/Vwfh7LR3brry2K7XdANsOyNvejsN7DfVo35fTb/ny5di7dy/eeecd9OrVS3JN36DnrelNDY4amrc9B1iJiCxZvUbEsfQ8JP15DeqKarg6OyDyrq4IC/KGTTuO2RnLLAJfXFwcNmzYgNdeew2PPvqoNt3d3R3AjV7azUpLSyW9Nnd3d735gBs9PGPKbAlRbNmgNQe82XYlUWq7Aetpu7qiBnGbjyM7twyC0PB/nyAAySeuwc/HBfMnh8LV2V5yT1OTW9qzz2HyMb5Vq1Zh7dq1WLRoEaKjoyXXGsfhdI27paenS8bpAgMDcf78eWg0Glk+oGH8ztgyiYhITqMREbf5OC7llQG4sTyh8e9LeWWI23wcGo15rtkwaeBbs2YNPvzwQ8ybNw/PPPOM7HqXLl3Qr18/JCYmSgJaamoqcnNzMWbMGG1aVFQU1Go19u3bJylj+/btCAgIQGBgoNFlEhGR3MmsQmTnluldiyiKQHZuGU5dKGzfihnIZtmyZctM8cIbNmxAbGws7r//fkyaNAk5OTnaP0VFRfDy8gIA9OjRAxs2bEBmZibc3Nxw5MgRvPHGG+jduzcWL16sncXp5+eHw4cPY/PmzfDw8IBarcaaNWvwyy+/4O2330ZAQID2tQ0tsyVEEaiqqjH6vg4d7AAA1dV1LX5tS8W2K6/tSm03YB1t3/5bFq4VVjSZp/Hx58Bgb22avrY7OdkrYwH71KlTcejQIZ3XunXrJum5/frrr4iPj8fZs2fh7OyM0aNHY9GiRbLxuPLycsTGxmL37t1Qq9UIDAzErFmzMHr0aNlrGFqmsbiA3Xhsu/LartR2A9bR9hX//RNnsoubzdfHzwOLnrpL+7W5LGA3WeCzVgx8xmPbldd2pbYbsI62r004hcNn85rcdk0QgMF9fPD8hL7aNHMJfCaf3EJERJYlom8XadATRXSokQYzUQQi+vq0b8UMxMBHRERG6d/TE34+LhAEYNIf25EYNwnfrnkKvXLPA2jo7fn5uKBfgHxrSnNgFuv4iIjIcqhUAhYO7wz/YaMk6eJfjyt7eDes42vPx5fGYOAjIiKjuMx5AV6bvpakbXlyEbwiIzChrw/6BXiabdADGPiIiMhAtseOwmPMfZK0+i53oOjwCUQ6OCBS921mh4GPiIiaVl8P96hI2J06IUku2boDtcNHmKhSLcfJLUREpJf9ju/hdYeHJOhVjx6D/NxSiwx6AHt8RESkg1Behs49u8nSi1KOoD7Qsvc0Zo+PiIgkHFfHyoJe5Zz5yM9TW3zQA9jjIyKiv6guX4LnPf1k6QVnL0DsZJ5r8lqCPT4iIqUTRbi8MFMW9NSr/x/y89RWFfQA9viIiBTN9ugf8HhgpCStvnsPFKUeBezt9dxl2Rj4iIiUqK4OHiOHwfbsGUlyScKPqI0YZqJKtQ8+6iQiUhiHhK3w6tpJEvSqH3ioYYmClQc9gD0+IiLFENSl6BzYXZZeeOBPaHr2MkGNTIM9PiIiBXCKXS4LehXzFyI/T62ooAewx0dEZNVU2RfhOWiALL0gPRuiu4cJamR67PEREVkjUYTLszGyoKf+YF3DEgWFBj2APT4iIqtje/ggPB6OkqTV+weg6PfDVrtEwRgMfERE1qK2Fh4jhsD2fKYkufj73agLjzBRpcwPH3USEVkBh+82w6ubpyToVY+biPzcUga9W7DHR0RkwYTSEnTu3UOWXnjoODT+ASaokfljj4+IyEI5LX9bFvQqFv2rYYkCg55e7PEREVkY1YUseA4Jk6UXZFyC6OZughpZFvb4iIgshSjCNeZvsqCnXvtJwxIFBj2DsMdHRGQBbA+kwmPCWElaXe8gFO9PBezsTFQry8TAR0Rkzmpq0GnoQNhcuihJLt65B3WDhpimThaOjzqJiMyUw+b/wsu3syToXZ/0WMMSBQa9FmOPj4jIzAjFRegc7C9LL/zjJDQ9/Nq/QlaGPT4iIjPi9PabsqBX8a8lDUsUGPRaBXt8RERmwCYrE53C75alF2RehujqZoIaWS/2+IiITEkU4fr3ybKgV7r+s4YlCgx6rY49PiIiE7FL+R3ujzwkSavr0xfFP/8G2PK/57bCd5aIqL1VV6NTWB/YXLsqSS7etQ91dw80UaWUg486iYjakerzz2Dv2lES9K4/PgX5eWoGvXbCHh8RUTsQCgvRuY984+jCo2nQ+HY3QY2Uiz0+IqI25vzmUlnQK//3Gw1LFBj02p1Je3w5OTlYv3490tLScPbsWVRWVuLzzz/HkCHSHQlGjhyJq1evyu5/9tlnsXDhQklaQUEBVqxYgf3796O6uhohISFYuHAh7r5bPk04MTERH3/8MS5cuAAPDw9MmDABc+bMgYODQ+s2lIgUySYzA52G3iNJE1Uq1OYVoErDB26mYtJ3Pjs7Gzt37kRISAjCw8Oxb98+vXkHDRokC3I+Pj6Sr6urqxETE4PKykosWbIE7u7u+OyzzxATE4NvvvkGISEh2rwJCQl4+eWX8dRTT+HVV1/F+fPnsXLlSly9ehVxcXGt21AiUhZRhNuTj8L+l58lyaUbvoTj36b89UWVCSpGgIkD36BBg5CamgoA2Lt3b5OBz9XVFWFh8vOnbrZlyxZkZGRg69at6Nu3LwBg8ODBePDBBxEbG4v169cDAOrr67FixQqMHDkSy5YtAwCEh4fDzs4OS5YsQUxMDEJDQ1uhhUSkNHa/JcH9sfGStNr+oSjZ/QtgawtHE9WLbjDpGJ9K1bovv3fvXgQFBWmDHgDY29tj3LhxSElJQXl5OQDg2LFjyM/Px6RJkyT3jx8/HnZ2dti9e3er1ouIFOD6dXTq11sW9Ip/2o8SrsszKxYzueXAgQO466670K9fP4wfPx5ff/01RFGU5MnIyEBQUJDs3uDgYNTX1yMrK0ubDwB69+4tyefo6Iju3btrrxMRGaLDl5/Bq4c3bPJytWnXn/xbwxKFMPn8AjIti/gV5L777kO/fv3QvXt3lJSU4Pvvv8cbb7yBixcv4tVXX9XmKykpgZubfHufxrTi4mJtvpvTb83beL0lBAFwczP+YYatrc1fr6+8ByFsu/LabjXtzs+HvW9XWXLN+QtQ+fpC12ZjVtP2FtDXdkFo53q078u1zNKlSyVfR0VFYcGCBfjiiy8wbdo0dOvWTXtNaOIdvPWavrxNlUFEBAA2Ly+Czar3JWl17y6HZv58E9WIDGURgU+XSZMmYceOHThx4oQ28Lm7u+vsrZWWlmqv3/x3SUkJPDw8ZHl9fX1bXC9RBEpbMFur8Tegltxr6dh25bXdktttc+4sOt07WJIm2tuj4OxFoGPHZmdrWnLbb5e+tnt6dmzXXp/FjPHdSqPRAJBOkAkMDER6eros77lz52BjY4OePXtq8wGQjeVVVVXh8uXLsrE/IiKIItweGy8LeqWf/RcFVwoagh5ZBIsNfAkJCVCpVOjfv782LSoqCunp6Thz5ow2raamBjt37kRERAQ6/vWNGRYWBi8vLyQkJEjK3LFjB2prazFmzJj2aQQRWQS7/fvg5eMG+9+StGm1d9+D/P8Vo+bBh01YM2oJkz/q3LVrFwDg5MmTAIDDhw+juLgYjo6OiIyMxI4dO/Dzzz8jMjISXbp0QWlpKb7//nvs3bsXM2fORNeuNwaWH3/8cXz11VeYPXs2FixYADc3N3z++efIy8vD++/feBZva2uLBQsWYPHixXjzzTcxduxY7QL2sWPHNrtekIgUoqoKnnf1gaqoSJJc/PNvqOvPtb6WShBvXRPQzoKDg3Wmd+vWDfv27cOxY8fw/vvvIzMzEyUlJbCzs0NwcDCmTJkiW4cHAPn5+Vi+fDmSkpK0W5YtWLAAAwfKdz1PSEjA+vXrtVuWjR8/HnPnzkWHDh1a3B6NRkRhYbnR9/G5P9uuJJbQ7g6fbYDLohclaVVTY1D+3urbKtcS2t5WmhrjU6nab5DP5IHP2jDwGY9tV17bzbndQl4eOvcLlKUXnjgHTZc7brt8c257WzOXwGexY3xERK3N+bWXZUGv/K13G05RaIWgR+bB5GN8RESmZnPmNDpFhkvSRCdnFKRlAs7OJqoVtRX2+IhIuTQauE18UBb0Sr/ajIKL/2PQs1Ls8RGRItnt2wP3Jx+TpNUODkfJ97uAVt5An8wLAx8RKUtlJTz7B0FVppYkF/2Sgvq+/UxUKWpP/LWGiBSjwycfwcu/iyToVcXMRH6emkFPQdjjIyKrp8rNgWd/+ZFlBSczIPr4mKBGZErs8RGRVev48nxZ0Ct7ZwXy89QMegrFHh8RWSWbtFPodP9QSZrGzR2Fx88CTk4mqhWZAwY+MisajYiTWYVITctBWWUtXJzsENG3C/r39GzXnR3Igmk0cB83BnZ/HJIkl3zzHWpHRpmoUmROGPjIbKgrahC3+Tiyc8sgCA1nGwoCcOhMHvx8XDB/cihcne1NXU0yY/Z7dsHtb5MlaTVDh6N06w4uUSAtfieQWajXiIjbfByX8soANAS9m/++lFeGuM3HodFwa1nSoaICnf18ZEGvKOkASrf/wKBHEvxuILNwLD0P2bll0LdluigC2bllOHWhsH0rRmbP8aMP4BVwB4SqGxsfVz77QsMShT4hJqwZmSs+6iSzkPTnNe3jTX0EAUhNy8WAXp3br2JktlQ5/4PnAPmxZgWnMiF6e5ugRmQp2OMjs6CuqG4y6AENQVFdUdM+FSKz1nHBXFnQK1vxfsMSBQY9agZ7fGQWXJ0dDOrxcXKLstmePA6PUfdK0jSenig8ehpwdDRRrcjSsMdHZiHyrq4G9fgi+nLBsSLV18N9TKQs6JVs2obCMxcY9MgoDHxkFsKCvOHn4wJBz1I9QQD8fFzQL8CzfStGJmf/40543eEBu2N/atNqRtyP/NxS1N4/yoQ1I0vFR51kFmxUAuZPDtW5jk8UgR7eDev4uIhdQcrL0TnYD0JtrSS56PfDqA+ST2ohMhQDH5kNV2d7LJk2EKcuFCI1LRfqihq4Otsjoq8P+gVw5xYlcfwwHh2XvSZJq3xhNirefNtENSJrwsBHZkWlEjCgV2cuWVAo1bWr8AzrI0svOJ0FsTO/J6h1cIyPiMyCy9x/yIJeWWx8wxIFBj1qRezxEZFJ2R47Co8x90nS6n26oOjwCaBDB9NUiqwaAx8RmUZ9PdxHDofdqROS5JLvElF7b6SJKkVK0OJHnTU1NcjNzUVNDXfSICLjCNu3wd6pgyToVY+KaliiwKBHbczowJeWlobo6GjcfffduO+++3DkyBEAQGFhIaZNm4aUlJRWryQRWQehvAxe3q6wm3LLKQopR6D+73fQu5CTqBUZFfjOnDmDv/3tb7h8+TImTpwouebp6Ynq6mps27atVStIRNbBcXUsOvfsJkmrnP1iwykKgb1NVCtSIqPG+FatWgVvb29s27YN1dXV+O677yTXw8PD8eOPP7ZqBYnIsqmuXIbn3X1l6TXXclBh62SCGpHSGdXjO3LkCJ544gk4OztD0PFIomvXrsjLy2u1yhGRBRNFuPzjGVnQU6/6EDXVtYAnt58j0zCqx1ddXQ0XFxe918vLy2+7QkRk+WyP/gGPB0ZK0up9u6Mo9Sjg4AAuUiBTMirw9ejRA2lpaXqvHzhwAIGBgbddKSKyUHV18Bg5DLZnz0iSS7b/gNqhw01UKSIpox51jhs3DgkJCZKZm42PPDds2IDffvtNNumFiJTB/vtt8OraSRL0qsc+2LBEgUGPzIhRPb4ZM2YgOTkZM2fORM+ePSEIAt555x0UFRWhoKAAQ4cOxdNPP91WdSUiMySoS9E5sLssvejAUdT35BMgMj9G9fjs7e3x6aef4pVXXoGDgwMcHBxw8eJFeHh4YNGiRfjoo4+gUnH7TyKlcIpdLgt6FfMXNixRYNAjMyWIYnPnXpMxNBoRhYXGT/Jxc2s4Qbq0tKq1q2T22HbLa7vqUjY8B/aXpRecuwjRo1Oz91tqu1sD2y5vu6dnx3Y9dozdMyIynCjC5bkYWdBTf7Cu4RQFA4Iekam1aJPq5ORkXLx4ESUlJbi1wygIAmbNmtUqlSMi82F7+CA8Ho6SpNX7+aMo+Q/A3t5EtSIynlGB7/z585g9ezYuXrwoC3iNjAl8OTk5WL9+PdLS0nD27FlUVlbi888/x5AhQ2R5ExMT8fHHH+PChQvw8PDAhAkTMGfOHDg4OEjyFRQUYMWKFdi/fz+qq6sREhKChQsX4u67725xmUSKVlsLjxFDYHs+U5Jc8v0u1IYPNVGliFrOqMD3+uuvIycnB6+++ioGDhwIV1fX23rx7Oxs7Ny5EyEhIQgPD8e+fft05ktISMDLL7+Mp556Cq+++irOnz+PlStX4urVq4iLi9Pmq66uRkxMDCorK7FkyRK4u7vjs88+Q0xMDL755huEhIQYXSaRkjls/RauL8yUpFU/PAHqDV9wQ2myWEYFvhMnTuC5557D1KlTW+XFBw0ahNTUVADA3r17dQa++vp6rFixAiNHjsSyZcsANOwJamdnhyVLliAmJgahoaEAgC1btiAjIwNbt25F374N2yQNHjwYDz74IGJjY7F+/XqjyyRSIqG0BJ1795ClFx48Bk1ATxPUiKj1GDW5xd3dHR4eHq334gYsfTh27Bjy8/MxadIkSfr48eNhZ2eH3bt3a9P27t2LoKAgbdADGpZgjBs3DikpKdot1Ywpk0hpnFa8Iwt6FQsXIz9PzaBHVsGowPfwww9j7969bVUXnTIyMgAAvXtLjy1xdHRE9+7dtdcb8wYFBcnKCA4ORn19PbKysowuk0gpVBcvwMvbFc4r3pGkF2RcQuXLr5qoVkStz6hHnfPnz8fcuXMxa9YsTJ06Fd26dYONjY0sX9euXVutgiUlJQAANzc32TU3Nzft9ca8+vIBQHFxsdFlGksQbqxVMYatrc1fr2/8vZbOHNterxFxLD0PSX9eg7qiGq7ODoi8qyvCgrxh04rrjcyi7aII2ymToUrYLkmu++xzaJ58Crc3kq+bWbTbRNh2edvbe7jYqMBna2uLwMBAfPLJJ3onogANB9a2Nl3HIOlK15fPmLxNlUHWr7S8Gm9tPISsq2oIAiCKDT+YySeuoWc3V/w7ZjDcOlrHzF8h+XfYjbxfkiYGBaP26J+AnZ2JakXUtowKfCtWrMDGjRvRp08f3HPPPTp7TK3N3d0dQEMv7dbxxdLSUvj6+kry6uqtlZaWSsoypkxjiWLLdmTgbg7m0XaNRsR/PvsDl/LKADR8njf/feGaGm+sP4gl0wa2yk4TJmt7TQ06DR0Im0sXJcnFO/egbtAQoLIOQF2bvbw5febtjW3XvXNLe/Y3jAp8CQkJiIqKwurVq9uqPjKNxxxlZGQgICBAm15VVYXLly/j/vvvl+RNT0+XlXHu3DnY2NigZ8+eRpdJynIyqxDZuWV6r4sikJ1bhlMXCjGgV+d2rFnrcfj2G7jOek6Sdv2RR1H20adcokCKYNTklqqqKgwbNqyt6qJTWFgYvLy8kJCQIEnfsWMHamtrMWbMGG1aVFQU0tPTJY9aa2pqsHPnTkRERKBjx45Gl0nKkpqW0+z//YIApKbltk+FWpFQXAQvb1dZ0Cs8fAJl6zYy6JFi2CxrXMhmgIMHD0IURYwYMaLVKrBr1y5kZmbi+PHjOHr0KHx9fVFUVISrV6/C398fKpUKHh4eWLduHYqLi9GhQwf8+uuvWL58OUaOHInp06drywoODsZPP/2ExMREeHl5IS8vD++++y7OnTuHlStXwtvbGwCMKtNYoghUVdUYfV+HDg3jKdXVbfd4yVyZU9t/+fMq8kuuN5vP1ckew/rfcduv115td3rnTbj/bbIkreKV11C6bSfEvx79tydz+szbG9sub7uTk327zq0w6nSGs2fPYubMmXj99ddbrVcUHBysM71bt26SCTQJCQlYv369dnux8ePHY+7cuejQoYPkvvz8fCxfvhxJSUnaLcsWLFiAgQMHyl7D0DKNwdMZjGdObV+bcAqHz+ahqZ8KQQAG9/HB8xP66s9koLZuu01WJjqFy7frK8i8DNG17cfo9TGnz7y9se2mP53BqMAXHR2NnJwcXL58GT4+PvD19ZUtQhcEAZ999lmrV9RSMPAZz5zafjyzAKu2nGg234tPDGiVMb42a7sowjX6STjs/lGSrP54I6onPtq6r9UC5vSZtze23fSBz6jJLVeuXAEA3HFHwyOea9eutX6NiEyof09P+Pm44FJemc5enyAAPbxd0C/As/0rZyC7lN/h/shDkrS6PiEo/vl3wLZFB7IQWRWjfgqaWrtHZA1UKgHzJ4cibvNxZOeWSdbxiWJD0Js/ObRdfzs1WE0NOg0Jg83VK5Lk4h9/Rt09g0xUKSLzw1//iG7h6myPJdMG4tSFQqSm5UJdUQNXZ3tE9PVBvwBPswx6Dt98Bde5/5CkXX9sMso+/JizNYluwcBHpINKJWBAr85mv1ZPKCpE5zsDZOmFR05B011+ugIRNRP4oqOjIQgCPvnkE9ja2iI6OrrZApU+uYWovTj/53U4xUvPjiz/9zJUzX3JRDUisgxNBr4rV65AEATtaeuNk1uIyHRsMjPQaeg9kjRREFCYeRmiS1tsKU1kXZoMfLdOZuHkFqK2o9GIOJlViNS0HJRV1sLFyQ4Rfbugf8+/xhVFEW5PPQb7fdKjwUo3fImacRNMVGsiy2PwGF9NTQ2OHz8OLy8v+Pv7t2GViJRHXVGjcybpoTN58PNxwau+pej6N+n6u9p+A1Dy034uUSAyksF7dapUKsTExODXX39ty/oQKU69RkTc5uM6T4SwravFm28+Lgt6xbt/Qck+rssjagmDA5+trS06d+4MIzZ6ISIDHEvPQ3aufMH8mJN7sG31E+hUUaxNuz7laeTnqVF31z0gopYx6nSGBx54AD/++CM0Gk1b1YdIcZL+vCZZaudWWYLE2EcwZ88Hknyrl29HWfzadq4dkfUx6jnJE088gYMHD2L69OmYNm0a/Pz84OjoKMvXtWvXVqsgkbVTV1Rre3uJsY/Irm8YEYNtAx9BHwcP2TUiMp5RgW/cuHHa5Q2HDh3Sm+/m8/CIqGkuzvYYmpGKfyX+n+zaE7P/i+v2jhCEhh1liOj2GRX4Zs2a1a5nJhFZu1J1Ff4VM0SWvqfvKKweO0f7tSgCEX192rNqRFbLqMA3Z86c5jMRkUGc3lgC5w9WydIfmbcF9TY3fjQFAD18zPtECCJLwrnQRO1MKC1B597yfTRXjZmNvf1Gy9K9PBzN90QIIgtk1KzOr776CjExMXqvz5gxA998883t1onIarmPHqEz6I1/abvOoAcA/l1cOL5H1IqMCnxbt26Fn5+f3uv+/v747rvvbrtSRNbG5nQavLxdYXfimCT9padXYPxL25u8t6yyti2rRqQ4RgW+7OxsBAUF6b0eGBiI7Ozs264UkTXx8nZFp/siJGn13XvgPx+nIPOO3k3ey9mcRK3PqMBXV1eHmpoavddrampQXV1925UisgYO27+Dl7f8tISCtPMoOnIKEX27yHZruRVncxK1PqMCn7+/P5KTk/Ve//3339GjBw+/JIWrr4eXtytcn5suSa6KnoH8PDVELy8AQP+enujZzVXvAemCAPhxNidRqzMq8D388MNITk7G+++/L+n51dbWYvXq1UhOTsa4ceNavZJElsJ5yWJ43SHfYSX/WhHKV74vSVOpBPw7ZjACujb0ChsDYOPfPbxdOJuTqA0IohG7TtfW1mLGjBk4fPgw3Nzc0LNnTwiCgPPnz6O0tBQDBw7Ehg0bYG+v3DEJjUZEYWG50fe5uTVs/VZaWtXaVTJ71tB2oagQne8MkKWr49eiesrTeu9zc3NEvUZEyrErSE3LhbqiBq7O9ojo64N+AZ5WG/Ss4TNvKbZd3nZPz47t+r1uVOADGoLfxo0bsWPHDu1EFn9/f4wfPx7R0dGws7Nrk4paCgY+41l62z2GD4Jt+jlZen6eutl7Lb3tLaXUdvU/DAEAAB41SURBVANsO2CBgY+axsBnPEttu83JE+g0argsvfjn31DXP9SgMiy17bdLqe0G2HbA9IGPO7eQ1dNoRJzMKkRqWg7KKmvh4mSHiL5d0L9nyx8l6pqtWRfYG8UpR263ukTUxhj4yKqpK2oQt/k4snPLIAgNywMEATh0Jg9+Pg2TR4xZJ+ewZRNc//msLL3g7AWInTj7ksgSMPCR1dJoRMRtPo5LeWUAoF0z1/j3pbwyxG0+jiXTBmp7fnp7hz3c4OMrD2yVz76Aiv9vebu0h4haBwMfWa2TWYXIzi3Te10UgezcMpy6UIgBvTrr7R3eveY/8Dn+o+z+/P8VAzY2bdkEImoDRq3jI7IkqWk5eheHNxIEIDUtV2fv0LWyFN+/9wgeviXoqdd+0jBjk0GPyCKxx0dWq6yy1qAtwdQVNbLe4cefPI8upbmy/D+nZmFAr86tXVUiakcMfGS1XJzstI8s9WncBLqxdzgw8xCWJrwtyzdn6vvI9vbH4LRcBj4iC8fAR1Yrom8XHDqT12Sexk2gdx+6jO/fe0R2PcvLH/Om/rXV2F+9QyKybAx8ZLX69/SEn48LLuWV6ez1CULDfpgRnyzHqPVrZdejn9uA4o6dJPl5RBCR5ePkFrJaKpWA+ZND0cPbBYB8E+ienRyw5pVRcLol6F3o7I/xL22XBD2ARwQRWQv2+MiquTrbY8m0gTh1oVCyCfTLi8bBvrhQlv+R+d+hXpDP1mzsHfKIICLLx8BHVk+lEjCgV2cM6NUZqgtZ8BwSJstTsfjfyHn+RfjqWMcnijwiiMiaWETgO3jwIKKjo3Ve++GHH9CrVy/t18nJyVi1ahXOnj0LZ2dnREVFYeHChXB1le6tWFFRgbi4OOzatQtqtRqBgYGYNWsWRo0a1aZtIdPRtb8mcOMUBVdAZ+/Q2o8IIlIaiwh8jRYuXIhBgwZJ0nx9fbX/PnjwIJ577jmMGjUKL774IvLy8rBy5Uqkp6fj66+/hkp1Y0hz9uzZOH36NBYuXAhfX19s27YNs2fPxtq1axEZGdlubaK25/jx/0PH116RpZck/IjaiGGStJt7h0RknSwq8AUEBCAsTP6YqtGKFSvQu3dvvP/++9og5+XlhRkzZmDXrl146KGHAABJSUlISUnBmjVrEBUVBQAIDw/H5cuX8e677zLwWZHmenlEpDxWM6szNzcXJ0+exMSJEyU9u2HDhsHHxwe7d+/Wpu3ZswcuLi6Sx5qCIGDSpEnIyspCZmZmu9adWp/7mEidQa/wz9MMekQKZ1GBb+nSpQgJCcE999yD559/HqdOndJeS09PBwD07t1bdl9QUBAyMjK0X2dkZCAwMFASIAEgODhYUhZZoPJyeHm7wu7Yn7JL+XlqaLr56riJiJTEIh51uri4YNq0aRg8eDDc3d1x/vx5rFu3Dk899RS+/PJLhIaGoqSkBADg5uYmu9/NzQ2nT5/Wfl1SUgJ/f3+d+Rqvt5Qg3Dhl2Bi2tjZ/1cH4ey1da7Xd3sFOZ3pN5XXAxgby7wzTU+rnrtR2A2w7IG97c5vJt3o92vflWiYkJAQhISHarwcOHIiRI0di3LhxiIuLw8aNG7XXBD3v4K3p+vI1d43Mj3D0COwiwmXp9TExqP/oYxPUiIjMmUUEPl28vLwwfPhw7Nu3DwDg7u4OQHdvrbS0VNITdHd315sP0N1rNJQoAqWlVUbf1/gbUEvutXS30/ZmJ6+Y+fup1M9dqe0G2HZA3nZPz47t2uuzqDG+W2k0Gu2/G8f2bh7La5Seni4Z+wsMDMT58+cl9zfmAxrGBMm8Oa6O0xn0Sr/YxMkrRNQkiw18+fn5SElJ0S5v6NKlC/r164fExERJQEtNTUVubi7GjBmjTYuKioJardb2Fhtt374dAQEBCAwMbJ9GUIt4ebui41uvy9Lz89SoGfugCWpERJbEIh51LliwAN27d0ffvn3h6uqKrKwsfPzxx7h+/Tpeeuklbb6FCxdi5syZeOmllzBlyhTk5uZi5cqVCA0NxQMPPKDNFxkZiSFDhuC1115DSUkJfH19sX37dhw5cgQffvihKZpIBvAYPgi26edk6YUnzkHT5Q4T1IiILJEgis2dUW1669atw86dO3H16lVUVVXB3d0dgwcPxj/+8Q/ZY8lff/0V8fHx2i3LRo8ejUWLFsnG7crLyxEbG4vdu3dLtiwbPXr0bdVVoxFRWFhu9H187q+/7UKZGp17yZchiCoVCnJaPgPXHCj1c1dquwG2HdA9xteeWwJaROCzJAx8xmuq7fomr+ReK8bJi8VITctBWWUtXJzsENG3C/r3tKw9NZX6uSu13QDbDpg+8FnEo05SHrtffob7lEmy9KoZz+LakncQ98VR2SkKh87kwc+n4RSFjo52OJlVaPGBkYhaHwMfmZ2mlihoNCLiPvsDl/LKAEB7snrj35fyyrDymz9ho1I1GRh5kjqRclnsrE6yPm6PjtMZ9NQfbdAuUTiZVYjs3DLoe0AvisCV/Apk5+oPjHGbj0Oj4RN+IqVij49MTxRh30F3D+zWNXmpaTnaXlwLXwrZuWU4daGQRw8RKRR7fGRSXt6uOoNe4R8ndS5EL6usbXHQayQIQGpa7u0VQkQWiz0+MglVzv/gOSBY57Wmdl5xcbK7rR4f0HCvuqKm5QUQkUVj4KN2p3fySk4JoGr6IURE3y44dCbvtl5fEMDJLUQKxsBHBtNoxNtaIuCw6Wu4znlBXu6AAag7fMSgDaX79/SEn48LLuXpn+DSHFEEIvr6tOxmIrJ4DHxkEHVFDeI2H2/xEoGmligYcy6ZSiVg/uRQnXURRaCHd0doRBFXCyp0BkZBAHp4u6BfgKfBr0lE1oWBj5ql0YiI23y8ybVzcZuPY8m0gbKen/uYSJ2noatXfYjqp/7eovq4OttjybSBOHWhEKlpuVBX1MDV2R4RfX3QL8AT5VW1TQTGhiDNRexEysXAR81qXDunj84lAqIILx/d5xq2xrFBKpWAAb0661yS0FxgZNAjUjYGPoUyZrzOkLVzjUsEBvTqrPexZuGxM9B07daazdCrqcBIRMrGwKdAxo7XGbJ2ThQBu4tZ8PKO0Hmdh8MSkblg4FOYlozXGbJ2LjH2EZ3p+bmlDVGViMhMcOcWhTFkr8vG8bpGEX276M0/a8+HOoNefdduDb08Bj0iMjPs8SmMseN1gP61c3p7eXysSURmjIFPYQwdr7t5S69b187pC3jlS/+DqtnzWqWet7tYnohIHwY+hTFkvE7Xll6uzvZY8ve74NOtk857WrOXd7uL5YmImsIxPoVparyuka4tvby8XXUGvcJDx1s16Bk6+Ybn6RFRSzHwKUzjeJ2+OSeCAPj53NjSy/bggaZPRPcPaNX6tWTyDRGRMRj4FKZxvK6HtwuAG5MuG/++eUsvL29XeIwfIysjP7e0zSawNE6+aQrP0yOi28ExPgVqbksvjycnwX7/Pp33Nga8tpp80pLJN0RExmDgUyh9W3o19VizkSGTTzo62rUoMLZ08g0RkaEY+AiA/oBX8eJCVL66VPu1IZNPVn7zJ2xUqhbNyjTkoFmep0dEt4NjfEpXU9NkL+/moAcYNvnkSn6F9jQHY2dlGjv5hojIWOzxWbHmxuH0nqJw5BQ03XvovGbIzi9N0XmE0U2aP2iW5+kR0e1h4LNSTY3DjS09h9mfvKLzvuZmaxoy+aQ5t26Jdiuep0dEbYmBzwo1NQ53u/trGjL5pDmGzMrkeXpE1FYY+KyQrhPT3/tqEYJyM3Tmf3lVElwSThk069KQySfN4axMIjIlBj4rdOs4nL5e3viXtjf8I7sYgGGzLvv6d4KXuyPyS6paXD/OyiQiU2Lgs0KN43D6At6n90Zj66BHdV7Lzm1YjrBs+mBZz69x3PB2gp4gNExQ4axMIjIVBj4r1FGoa76X14Qr+RU4cb4QYb1vjK9pxw1veYRqKM7KJCJzwcBnZby8XbFUR3r0cxtQ3FH3kUK67D50SRv4NBoR3ydfkI0bGiK4uztUKoGzMonIbDDwWQBD9sW0S02G+8QHdd5vSC/vVv8rqgQgXRZhLEEA3F0c8PyEvkbfS0TUVhj4zFxT6/G83B3h6+WMN/8xXOe9LQl4WqIoWxbRgiK4mTQRmR0GPjNRrxFxLD0Pew9d0vbqwkN8sP23C7icXw5Avv3XwrUvIuTaGVlZ5Q5OeGrW17dVnzs8nXUuizAGly0QkTli4DMD6ooavPXFH8i6qpb16vS5nckrhrg39A6kpuXcVhlctkBE5kjRga+iogJxcXHYtWsX1Go1AgMDMWvWLIwaNard6qDRiIjddAyX8nT36m6lL+D938ML8Xuw7keeLfHToctwdLBp8f1ctkBE5krRpzPMnj0biYmJmDdvHj766CMEBgZi9uzZSEpKarc6fLQpWRv0mtKhpqrJXp4xQa+za4dm81zKK0dZVa3BZd6KyxaIyFwptseXlJSElJQUrFmzBlFRUQCA8PBwXL58Ge+++y4iIyPbvA4z3tV9yvmt9AW8p/7xBcodXYx6TacOtuji6YQC9fUm8wlCw0L4lvB2d8S8xwdwfI+IzJJie3x79uyBi4uL5LGmIAiYNGkSsrKykJmZ2aav/32yfFLKre49+1uTvTxjgx4AzHyoDzKvljabTxQbJty0RH5pFVZtOaH3zD0iIlNSbI8vIyMDgYGBUKmksT84OBgAkJ6ejsDAwDZ7/e2//a/J6609eaVxzA0CcL2m3qB7HO1tUFVdZ/RJDM2duUdEZEqKDXwlJSXw9/eXpbu5uWmvt4QgAG5uji2u19ZVj8Ouvk6WXmNjh8fmfdvicgO6uuLfMYPxSeJpg48VGjWoBzb/rPtEh+YIAvBHegHuvbt7s3ltbRsm0dzO+2aplNp2pbYbYNsBeduFdp4KoNjABzQ82mzJtbbSVksUJo/qjcdH9oaNSoC6otqgoOfoYINH7wvEH2dzceGaukW9vtLy6pZVmIioDSk28Lm7u+vs1ZWWNox/Nfb8jCWKQGmpcacX2NdW47v4KbL0tfc/i513PdyiegA3Hm+OuccX5WUNk1kc7W0M6vH18euEyopqzH10gGznGENf28nB1qD3ovG3P2PfN2ug1LYrtd0A2w7I2+7p2bFde32KDXyBgYH46aefoNFoJON86enpAICgoKB2q4uuoHc7vbymTkIw9CDZEaF3AGjYeWXJtIE4daEQqWm5UFfUoF4jIv1y04+CuXidiMyVYgNfVFQUtmzZgn379mH06NHa9O3btyMgIKBNJ7YAwCP33qGd4DLxxe+Q8P5jAIAps75CpYNzi8p0dLCBn48L3Do66D0JoX9PT/j5uOBSXpnO3puuhecqlYABvTprJ6poNCL+89kfRpVBRGQuFBv4IiMjMWTIELz22msoKSmBr68vtm/fjiNHjuDDDz9s89efMKyPNvBpVDZ6e3iNvTdvD0d4uzvi3OUS1NZpZPmaOzm9kUolYP7kUJ0bXxt6Xl5rlEFEZCqCKBo7bcF6lJeXIzY2Frt375ZsWXZzD9BYGo2IwsLmd2Jp9H3yGZ1LG7ycgM5eHrJz7DQaUfLYsaXn3LVGOa1VF455KK/tSm03wLYDusf42vMXZUUHvrZgbOBrxB8Gtl1JlNpugG0HTB/4FLtzCxERKRMDHxERKQoDHxERKQoDHxERKQont7QyURSN3t4LuLFXnRI/DbZdeW1XarsBth2Qt10Q2nebSAY+IiJSFD7qJCIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgIyIiRWHgM7GKigq89dZbGD58OAYMGIBHH30UP//8s6mr1azU1FQsXrwYY8eORWhoKEaMGIHZs2fj3LlzsrzJycmYPHkyBgwYgIiICCxduhRqtVqWz5j3wtAy20N8fDyCg4MxceJE2TVrbfvBgwcxY8YMDBw4EKGhoXjooYewadMmSZ7ExERMmDAB/fv3x4gRI7By5UpUV1fLyiooKMArr7yCIUOGICwsDE8//TSOHj2q83UNLbMtnD59Gv/85z8xfPhwhIWF4aGHHsK6detQU1MjyWfpn3lOTg7eeustPPXUU7jrrrsQHByMgwcP6sxrys/YmDJlRDKpmJgYcfDgweLmzZvFlJQUcdGiReKdd94p7t+/39RVa9KcOXPEqVOnil9//bV48OBBcefOneKkSZPEfv36iX/++ac234EDB8SQkBBxzpw5YnJysrht2zZx2LBh4pQpU8T6+npJmYa+F8aU2dbS09PF/v37i0OHDhUnTJjQ4npaUtu3bt0q9unTR1y6dKmYlJQkpqSkiF9++aX4xRdfaPNs375dDAoKEl9//XUxNTVV/PLLL8WwsDDxxRdflJR1/fp18eGHHxbvv/9+MTExUfztt9/EZ555Ruzfv7+YlpYmyWtomW0hMzNT7N+/vzhhwgRx586dYkpKihgbGyveeeed4qJFi7T5rOEzP3DggBgeHi7OmDFDfOGFF8SgoCDxwIEDsnym/IyNKVMXBj4T2r9/vxgUFCT+9NNP2jSNRiM++eST4gMPPGDCmjWvoKBAllZaWioOHDhQnD17tjbtscceEydOnCj5Af3999/FoKAgcefOndo0Y94LQ8tsa/X19eITTzwhvvnmm+Lf//53WeCzxrZfu3ZNHDBggLhu3Tq9eerq6sRhw4aJL7zwgiR906ZNYlBQkHjs2DFt2pdffikGBQWJp06d0qZVV1eLI0eOFGfOnNmiMtvC6tWrxaCgIDE7O1uSvnDhQjEkJESsqakRRdE6PvObX2fPnj06A5+pP2NDy9SHjzpNaM+ePXBxccGoUaO0aYIgYNKkScjKykJmZqYJa9c0T09PWZqrqyv8/PyQk5MDAMjNzcXJkycxceJEqFQ3vtWGDRsGHx8f7N69W5tm6HthTJltbePGjcjJycH8+fNl16y17Vu2bAEATJ06VW+eY8eOIT8/H5MmTZKkjx8/HnZ2dpJ67t27F0FBQejbt682zd7eHuPGjUNKSgrKy8uNLrMt2NraAgA6duwoSXdxcYGtrS1sbGys5jO/+XX0MfVnbGiZetvYbAupzWRkZCAwMFD2jRYcHAwASE9PN0W1WqyoqAgZGRno3bs3gBv1b/z6ZkFBQcjIyNB+beh7YUyZbeny5ctYvXo1li5dKvvPELDeth8+fBi9evXCTz/9hLFjx6JPnz7acZjGsa7GetxaT0dHR3Tv3l3W9qCgINnrBAcHo76+HllZWUaX2RYmTpwId3d3LFu2DJcvX0Z5eTn27t2Lbdu2Yfr06VCpVFb7meti6s/Y0DL1YeAzoZKSEri5ucnSG9NKSkrau0otJooilixZAo1Gg5kzZwK4UX99bby5fYa+F8aU2VZEUcS///1vDB8+HKNHj9aZx1rbnpeXh4sXL+Ktt95CdHQ0Nm7ciMceewyffvop/vWvfxldz+baXlxcbHSZbaFr167YtGkTzp8/j9GjR+Oee+7BrFmzEB0djRdffNHoOlrSZ66LqT9jQ8vUx7bJq9TmBEFo0TVzs3z5cuzduxfvvPMOevXqJbmmrx23phvzXhhaZlvYvHkzTp06hR9++KHZvNbWdlEUUVFRgdjYWDz88MMAgCFDhuD69evYsGED5s6da3Q9LaHtV69exQsvvAAvLy988MEHcHFxweHDh/HRRx9BEARt8DOmjpbQ7uaYsq23838nA58Jubu76/yNrbS0FIDu33zMUVxcHDZs2IDXXnsNjz76qDbd3d0dgO6ea2lpqaR9hr4XxpTZFoqKirBixQo8//zzcHR01E4pr6urg0ajgVqthoODg1W2/eY6DB8+XJI+YsQIbNiwAWlpaZJ6enh4yOrp6+srKa+ptjeWZUyZbeG9995DRUUFtm/fjg4dOgBoCPgA8MEHH+Dxxx+32s9cF1N/xoaWqQ8fdZpQYGAgzp8/D41GI0lvfK6v6xm2uVm1ahXWrl2LRYsWITo6WnKt8Vm9rnGI9PR0ybN8Q98LY8psC7m5uSgrK8N7772HQYMGaf8cPXoU6enpGDRoEOLj462y7TfXRR+VSoXAwEAA8npWVVXh8uXLsrbrGss+d+4cbGxs0LNnT20+Q8tsC6dPn0ZgYKA26DXq168fNBoNsrKyrPYz18XUn7GhZerDwGdCUVFRUKvV2LdvnyR9+/btCAgI0H4jmKs1a9bgww8/xLx58/DMM8/Irnfp0gX9+vVDYmKi5Ac8NTUVubm5GDNmjDbN0PfCmDLbQo8ePfD555/L/tx5553aa1OmTLHKtjfWFQCSkpIk6UlJSRAEAf3790dYWBi8vLyQkJAgybNjxw7U1tbK2p6eno4zZ85o02pqarBz505ERERoJw4ZU2Zb8Pb2RkZGBqqqqiTpf/75JwDAx8fHaj9zXUz9GRtapj42y5YtW2Z0q6lV+Pn54fDhw9i8eTM8PDygVquxZs0a/PLLL3j77bcREBBg6irqtWHDBsTGxuL+++/HpEmTkJOTo/1TVFQELy8vAA2BYsOGDcjMzISbmxuOHDmCN954A71798bixYu1s9qMeS8MLbMt2Nvbw9fXV/bnhx9+gEajwcKFC7WPn6yt7Y11PXHiBDZt2gQHBwdUVlbi22+/xaeffoopU6ZgwoQJUKlU8PDwwLp161BcXIwOHTrg119/xfLlyzFy5EhMnz5dW15wcDB++uknJCYmwsvLC3l5eXj33Xdx7tw5rFy5Et7e3gBgVJltwcXFBd9++y0OHz6Mjh07Ii8vD1u2bMEnn3yC8PBw7YQua/nMd+3ahczMTBw/fhxHjx6Fr68vioqKcPXqVfj7+5v8Mza0TH0EURTFVn/XyGDl5eWIjY3F7t27oVarERgYiFmzZumdLWgupk6dikOHDum81q1bN8lvsr/++ivi4+Nx9uxZODs7Y/To0Vi0aJFsfMKY98LQMtvL1KlToVarZb+tWmPbKysrER8fjx07dqC4uBh33HEHnnjiCTzzzDOS/4QTEhKwfv16XLhwAR4eHhg/fjzmzp0re1yYn5+P5cuXIykpCdXV1QgJCcGCBQswcOBA2WsbWmZbSElJwbp165Ceno7Kykp069YNDz30EKZPnw4nJydtPmv4zBuXVdzq1p9tU37GxpR5KwY+IiJSFI7xERGRojDwERGRojDwERGRojDwERGRojDwERGRojDwERGRojDwEZm5gwcPIjg4GFu3bgUAXLlyBcHBwYiPj29xmVOnTsXIkSNbq4pEFoWBj4iadObMGcTHx+PKlSumrgpRq+DpDEQWplu3bjhx4gRsbGza5fXOnDmDNWvWYPDgwW1+CgJRe2DgI7IwgiDAwcHB1NUgslh81ElkYfSN8VVVVeGdd97B8OHDERoaismTJyM1NRWLFy/Wu/dibm4uXnrpJQwaNAhhYWGYOXMmLly4oL0eHx+vPVk9OjoawcHBCA4OxuLFi9uugURtjD0+Iisxb948JCUlYfTo0Rg6dCiuXLmCWbNm6X08WVlZib///e8IDQ3F/PnzceXKFXz++ef45z//iR07dsDGxgZRUVHIz8/Hpk2b8MILL2jPOevRo0d7No2oVTHwEVmBpKQkJCUl4YknnsBbb72lTQ8PD8dzzz2n857i4mLMnDkTzz77rDatU6dOWLFiBVJSUnDvvffizjvvRFhYGDZt2oShQ4dqTx0nsmR81ElkBRqPirn1XLrIyEj06tVL5z0qlQrR0dGStPDwcABAdnZ2G9SSyDww8BFZgStXrkClUul8BKnvQGNvb2/ZJBl3d3cAQElJSetXkshMMPARWRFBEAzO29RyCB7TSdaMgY/ICnTr1g0ajUbnI8qbZ2m2hDHBlMgSMPARWYHG7cc2btwoSU9KSsL58+dvq2wnJycAQGlp6W2VQ2QuOKuTyApERkZi+PDh2Lx5M4qLixEREYErV65g8+bNCA4Oxrlz51pcdv/+/aFSqbB27VqUlpbCyckJvr6+CA0NbcUWELUf9viIrIAgCIiPj0d0dDSOHj2K//u//8Mff/yBNWvWwN/fHx06dGhx2V27dsXbb7+N69ev44033sBLL72E//73v61Ye6L2JYgcxSayauPHj0dtbS127dpl6qoQmQX2+IisxPXr12Vp+/fvR3p6OoYNG2aCGhGZJ47xEVmJDz74AKdPn8aQIUPg4uKCM2fOYOvWrXB3d5fszkKkdHzUSWQlkpKSsG7dOmRmZqK8vBxubm4IDw/HvHnz4OfnZ+rqEZkNBj4iIlIUjvEREZGiMPAREZGiMPAREZGiMPAREZGiMPAREZGiMPAREZGi/P8SJlsjzfOilAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "crime2=list_crime_division(10)\n",
    "light2= list_light_division(10)\n",
    "crime_vs_light2=pd.DataFrame(list(zip(crime2,light2)),columns=['crime','light'])\n",
    "crime_vs_light2= crime_vs_light2.loc[(crime_vs_light2['crime']!=0) | (crime_vs_light2['light']!=0)]\n",
    "crime_vs_light2\n",
    "#pd.plotting.scatter_matrix(crime_vs_light2) \n",
    "\n",
    "func,axis = plt.subplots()\n",
    "\n",
    "plt.ylabel('crime')\n",
    "plt.xlabel('light')\n",
    "\n",
    "x=np.array(light2)\n",
    "y=np.array(crime2)\n",
    "\n",
    "linreg= LinearRegression()\n",
    "x=x.reshape(-1,1)\n",
    "linreg.fit(x,y)\n",
    "y_pred=linreg.predict(x)\n",
    "plt.scatter(x,y)\n",
    "plt.plot(x,y_pred,color='red')\n",
    "\n",
    "#outcome_1, predictors_1 = patsy.dmatrices('light ~ crime', crime_vs_light2)\n",
    "#mod_1 = sm.OLS(outcome_1, predictors_1)\n",
    "#res_1 = mod_1.fit()\n",
    "#print(res_1.summary())\n",
    "\n",
    "#crime_vs_light2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 168,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7f2cb1983080>]"
      ]
     },
     "execution_count": 168,
     "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": [
    "crime2=list_crime_division(50)\n",
    "light2= list_light_division(50)\n",
    "crime_vs_light2=pd.DataFrame(list(zip(crime2,light2)),columns=['crime','light'])\n",
    "crime_vs_light2= crime_vs_light2.loc[(crime_vs_light2['crime']!=0) | (crime_vs_light2['light']!=0)]\n",
    "crime_vs_light2\n",
    "#pd.plotting.scatter_matrix(crime_vs_light2) \n",
    "\n",
    "func,axis = plt.subplots()\n",
    "\n",
    "plt.ylabel('crime')\n",
    "plt.xlabel('light')\n",
    "\n",
    "x=np.array(light2)\n",
    "y=np.array(crime2)\n",
    "\n",
    "linreg= LinearRegression()\n",
    "x=x.reshape(-1,1)\n",
    "linreg.fit(x,y)\n",
    "y_pred=linreg.predict(x)\n",
    "plt.scatter(x,y)\n",
    "plt.plot(x,y_pred,color='red')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "As we increase our subregions the linear relations disappears."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "crime1=list_crime_division(20)\n",
    "light1= list_light_division(20)\n",
    "crime_vs_light1=pd.DataFrame(list(zip(crime1,light1)),columns=['crime','light'])\n",
    "crime_vs_light1= crime_vs_light1.loc[(crime_vs_light1['crime']!=0) | (crime_vs_light1['light']!=0)]\n",
    "#pd.plotting.scatter_matrix(crime_vs_light1) \n",
    "\n",
    "\n",
    "#st_cr, p_cr = normaltest(crime1)\n",
    "#st_lt, p_lt = normaltest(light1)\n",
    "df=crime_vs_light1['crime']\n",
    "\n",
    "xs = np.arange(df.min(), df.max(), 0.1)\n",
    "fit = stats.norm.pdf(xs, np.mean(df), np.std(df))\n",
    "# most easily done using matplotlib\n",
    "plt.plot(xs, fit, label = 'Normal Dist.', lw = 4)\n",
    "plt.hist(df, density = True, label = 'Actual Data');\n",
    "plt.title('lights per block Data')\n",
    "plt.legend();\n",
    "\n",
    "#we can see that since number of crimes are not normally distrubuted over the divided regions"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Ethics & Privacy"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "-Permission to the dataset:\n",
    "\t\n",
    "    The dataset we will use is allowed for public access, meaning no necessary permission is needed. The dataset is from https://data.sandiego.gov/ which is an open source dataset website. \n",
    "\n",
    "-Privacy Concern:\n",
    "\t\n",
    "    As a result of the number of streetlights and crimes about statistics in San Diego, we are confident that the datasets we used are unbiased. No names are included in the dataset that identify victims or accused persons. \n",
    "\n",
    "-Other Impact:\n",
    "\t\n",
    "    From the dataset it reveals the exact streets and cities that crimes happened, so this will certainly affect the property prices and popularity in those certain areas. Since these crimes are reported by the San Diego Police, there is potential for human bias in the crimes that are entered into the database and those that are not. There may be an uneven distribution of reported crimes in low-income areas or areas with greater police presence. It may be important for us to compare the distribution of crimes with the distribution of income as well.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Conclusion & Discussion"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Based on the dataset we have, it shows in the United States, authorities in many cities use light as a crime prevention measure. We first assume that it would be expected that crime rates would be lower in the brightest communities. However we found that the corelations between the two were very minimal. As the two dataset contain too much unhelpful data. During data cleaning, we have to clean the data which would not be affected by streetlight, like \"domestic voilence\", \"shopliftting\", and crime that happen during daytime. the size of our data were limited due to the limited quota access of the geocoder, where we have to convert address to longitudes and latitudes. where we had to cut down our size of data and focus mainly in city of San Diego.\n",
    "To keep the streetlight and crime ratio the same across our data, We try to cut the areas by different squares that each has the same size. From the 10*10 cut, meaning we divide the area in our dataset to 100 squares, we get the graph that is mostly linear, it seems there are positive relationships between the streetlights and the numbers of crimes. this is not what we expect.  However, as we change the cut to 20*20 or even up to 50*50. We got the graph that shows no relation between crime cases and existence of streetlights. This can happen because of serval factors: some data are skewed towards the police station, meaning that more crimes happens around our in the police station office.(about 5% of crimes where assult charges\n",
    "in the police stations.). Income distributions in different neiborhoods. Crime happens more often in some neibhorhoods then \n",
    "another. Despite having the same density of streetlights.\n",
    "\n",
    "Our hypothesis for this project is that the crime rate will be lower if there are more streetlights in the city. Because our initial research involved looking at cities like New York and Los Angeles, these may have been poor models in terms of comparison to San Diego. While the latter two are the product of city planning and have conventional “downtowns” – connected through transit hubs, the urban geography of San Diego differs in that it is composed of several disparate neighborhoods versus one mass “borough” (see: Manhattan). This may allow for crime differently than in these model cities. However, through the analysis of datasets of streetlights and crime in San Diego, we find that the result is somewhat different from our initial hypothesis. From our analysis of above, we might think that the relation between streetlights and crime is quite obvious, and consider the simple fact that crime in the United States is concentrated in cities, which are the brightest places on the map. If the relation between the streetlights and crime were as straightforward as we think, crime rates would fall in the brightest cities. Although our project cannot prove the relation between more streetlights and less crime, the fact is that people believe brighter streetlights at night make cities safer.For the final conclusion there are no correlations between streetlights and crime rates giving the data sets that we use. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}