Wieker-Alex-finalproject-part3.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Chicago Crime is not Alone by Alex Wieker"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Readme\n",
    "\n",
    "The two datasets that I use for my visualizations were too large to upload to Github, so the files that I am reading into my visualizations are already processed and on GitHub. However, I recognize that users might desire to work with the original files, so I have uploaded them to box.\n",
    "\n",
    "I would also like users to be aware that sometimes bqplot is finicky and causes errors when interacting with a visualization. If this happens simply re-run the cell. \n",
    "\n",
    "GitHub: https://github.com/senderrr/.github.io\n",
    "\n",
    "Crimes - 2001 to present (Chicago): https://uofi.app.box.com/s/r6842mh82l2o7s1gvp9x2rwi2ea1v31m\n",
    "\n",
    "Crime Data from 2010 to Present (Los Angeles):\n",
    "https://uofi.app.box.com/file/438183687378"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import bqplot\n",
    "import ipywidgets\n",
    "import matplotlib\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # read in the dataset\n",
    "# crimes_df = pd.read_csv(\"Crimes_2001_to_present.csv\")\n",
    "\n",
    "# # rename crime types that  will interfer with visualizations\n",
    "# crimes_df['Primary Type'] = crimes_df['Primary Type']\\\n",
    "# .replace(['MOTOR VEHICLE THEFT', 'CONCEALED CARRY LICENSE VIOLATION', \n",
    "#                               'INTERFERENCE WITH PUBLIC OFFICER', 'OTHER NARCOTIC VIOLATION',\n",
    "#                               'PUBLIC PEACE VIOLATION', 'NON-CRIMINAL (SUBJECT SPECIFIED)', \n",
    "#                              'OFFENSE INVOLVING CHILDREN', 'CRIM SEXUAL ASSAULT', \n",
    "#                               'LIQUOR LAW VIOLATION', 'WEAPONS VIOLATION'],\n",
    "#                              ['VEHICLE THEFT', 'CONCEALED CARRY VIO', 'INTERFER WITH POLICE', \n",
    "#                              'OTHER NARCOTIC VIO', 'PUBLIC PEACE VIO', 'NON-CRIMINAL SUBJ', \n",
    "#                              'CHILDREN', 'SEXUAL ASSAULT', 'LIQUOR LAW VIO', 'WEAPONS VIO'])\n",
    "\n",
    "# # remove domestic violence because there is only one reported domestic violence, \n",
    "# # which suggests that it is an outlier. Additionally, there is another column devoted to domestic violence \n",
    "# crimes_df = crimes_df[crimes_df['Primary Type'] != 'DOMESTIC VIOLENCE']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # take a quick view of the dataset\n",
    "# crimes_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # using this variable for later, which I will use as my first plot's x and y-values. \n",
    "\n",
    "# # group by the primary type column by the number of crimes of that type in descending order\n",
    "# crime_type_size_df= crimes_df.groupby('Primary Type')\\\n",
    "# .size().sort_values(ascending=False).reset_index().rename(columns={0: 'size'})\n",
    "\n",
    "# # remove other offenses because these are crimes that are lumped together as one crime \n",
    "# crime_type_size_df = crime_type_size_df[crime_type_size_df['Primary Type'] != 'OTHER OFFENSE']\n",
    "\n",
    "# # get only the most ten common crime type values\n",
    "# crime_type_size_df = crime_type_size_df.head(10).reset_index(drop=True)\n",
    "\n",
    "# crime_type_size_df.to_csv('crime_type_size_df.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # subset the primary type column in the crimes_df for the ten most common crime types (for figure two)\n",
    "# crime_type_year_df = crimes_df\n",
    "# crime_type_year_df['Primary Type'] = crime_type_year_df['Primary Type']\n",
    "# crime_type_year_df = crime_type_year_df[(crime_type_year_df['Primary Type'] == 'THEFT') \\\n",
    "#             | (crime_type_year_df['Primary Type'] == 'BATTERY')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'CRIMINAL DAMAGE')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'NARCOTICS')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'ASSAULT')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'VEHICLE THEFT')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'DECEPTIVE PRACTICE')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'ROBBERY')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'BURGLARY')\n",
    "#             | (crime_type_year_df['Primary Type'] == 'CRIMINAL TRESPASS')]\n",
    "\n",
    " \n",
    "# # create new dataframe for plot two that will display crime frequency for each primary type    \n",
    "# crime_type_year_df = crime_type_year_df.groupby(['Primary Type', 'Year']) \\\n",
    "#             .size().sort_values(ascending=False).reset_index().rename(columns={0: 'size'})\n",
    "\n",
    "# crime_type_year_df = crime_type_year_df[crime_type_year_df['Year'] < 2019]\n",
    "\n",
    "# crime_type_year_df.to_csv('crime_type_year_df.csv')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Primary Type</th>\n",
       "      <th>size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>THEFT</td>\n",
       "      <td>1439000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>BATTERY</td>\n",
       "      <td>1247989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>CRIMINAL DAMAGE</td>\n",
       "      <td>779869</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NARCOTICS</td>\n",
       "      <td>715960</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>ASSAULT</td>\n",
       "      <td>425066</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      Primary Type     size\n",
       "0            THEFT  1439000\n",
       "1          BATTERY  1247989\n",
       "2  CRIMINAL DAMAGE   779869\n",
       "3        NARCOTICS   715960\n",
       "4          ASSAULT   425066"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read in aggregate csv file\n",
    "crime_type_size_df = pd.read_csv(\"crime_type_size_df.csv\").drop(columns='Unnamed: 0')\n",
    "crime_type_size_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>Primary Type</th>\n",
       "      <th>Year</th>\n",
       "      <th>size</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>THEFT</td>\n",
       "      <td>2001</td>\n",
       "      <td>99264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>THEFT</td>\n",
       "      <td>2003</td>\n",
       "      <td>98875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>THEFT</td>\n",
       "      <td>2002</td>\n",
       "      <td>98327</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>THEFT</td>\n",
       "      <td>2004</td>\n",
       "      <td>95463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>BATTERY</td>\n",
       "      <td>2002</td>\n",
       "      <td>94153</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Primary Type  Year   size\n",
       "0        THEFT  2001  99264\n",
       "1        THEFT  2003  98875\n",
       "2        THEFT  2002  98327\n",
       "3        THEFT  2004  95463\n",
       "4      BATTERY  2002  94153"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read in aggregate csv file\n",
    "crime_type_year_df = pd.read_csv(\"crime_type_year_df.csv\").drop(columns='Unnamed: 0')\n",
    "crime_type_year_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "078b58ce1c8f43e7b130f6bbb4e0e264",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(Label(value=''), HBox(children=(Figure(axes=[Axis(label='Type of Crime', offset={'scale': Ordin…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# create plot one x and y values\n",
    "x = crime_type_size_df['Primary Type']\n",
    "y = crime_type_size_df['size']\n",
    "\n",
    "# initiate plot one scales\n",
    "x_sc = bqplot.OrdinalScale()\n",
    "y_sc = bqplot.LinearScale()\n",
    "\n",
    "\n",
    "bar = bqplot.Bars(x=x, y = y/1000, # set y-values to per thousand, so values are more clear.\n",
    "                scales={'x': x_sc, 'y': y_sc}, \n",
    "                interactions={'click':'select'},\n",
    "                selected_style={'fill': 'orange'})\n",
    "\n",
    "ax_x = bqplot.Axis(scale=x_sc, \n",
    "                   tick_style= {'font-size': '10px'}, # set font-size to ten for increased legibility\n",
    "                   tick_rotate = -700,  # rotate x-axis values\n",
    "                   offset={'scale': x_sc, 'value': '50'}, # offset x-axis values\n",
    "                   label= 'Type of Crime')  # set x-axis label\n",
    "\n",
    "ax_y = bqplot.Axis(scale=y_sc,\n",
    "                   orientation='vertical', \n",
    "                   label='Total Crimes Reported per Thousand Crimes', # set y-axis label\n",
    "                   label_offset='-50') # y-axis label offset\n",
    "\n",
    "# initiate a label above the plot\n",
    "mySelectedLabel = ipywidgets.Label()\n",
    "\n",
    "# initiate plot year label range\n",
    "year_range = bqplot.Label(x=[0.75],y=[0.10], font_size=52, font_weight='bolder', \n",
    "                         colors=['blue'], \n",
    "                          text=[('Year Range: 2001 to 2019')], \n",
    "                          x_offset = -125,\n",
    "                          y_offset = -300)\n",
    "\n",
    "fig = bqplot.Figure(marks=[bar, year_range], axes=[ax_x, ax_y], padding_x=0.025, padding_y=0.025,\n",
    "                    title=\"Ten Most Common Reported Chicagoan Crimes\")\n",
    "\n",
    "# initiate plot two scales\n",
    "x_sc2 = bqplot.LinearScale()\n",
    "y_sc2 = bqplot.LinearScale()\n",
    "\n",
    "\n",
    "x_ax2 = bqplot.Axis(scale=x_sc2, label='Year', # set x-axis label\n",
    "                    tick_values=crime_type_year_df['Year'], # set tick values\n",
    "                    offset={'scale': x_sc, 'value': '200'}, # offset x-values (scale has to be x_sc to work)\n",
    "                    tick_rotate = 400) # rotate x-values\n",
    " \n",
    "y_ax2 = bqplot.Axis(scale=y_sc2, \n",
    "                    label='Total Crimes Reported', # set label\n",
    "                    orientation = 'vertical', \n",
    "                    label_offset='-50')  # set label off set\n",
    "\n",
    "# initiate the start plot two subset--the get_tot_crimes_types function will change subset by clicking\n",
    "mask = 'THEFT'\n",
    "\n",
    "# create a mask that updates plot two based upon selection. The default mask value is theft as the starting point. \n",
    "# Instructor Naiman helped me set up the mask. \n",
    "x_mask = crime_type_year_df['Year']\n",
    "x_mask = x_mask[crime_type_year_df['Primary Type'] == mask]\n",
    "\n",
    "# create a mask similar to x_mask except for the y values\n",
    "y_mask = crime_type_year_df['size']\n",
    "y_mask = y_mask[crime_type_year_df['Primary Type'] == mask]\n",
    "\n",
    "line_plot = bqplot.marks.Lines(x=sorted(x_mask), y = y_mask,marker='circle', stroke_width=1, scales={'x':x_sc2,'y':y_sc2})\n",
    "\n",
    "fig2 = bqplot.Figure(marks=[line_plot], axes=[x_ax2,y_ax2], \n",
    "                     title=\"Total Crimes by Crime Type and Year\")\n",
    "\n",
    "# create change function that subsets first plot by click\n",
    "def get_tot_crime_types(change):\n",
    "    mask = x[change['owner'].selected[0]]\n",
    "\n",
    "    x_mask = crime_type_year_df['Year']\n",
    "    x_mask = sorted(x_mask[crime_type_year_df['Primary Type'] == mask])\n",
    "\n",
    "    y_mask = crime_type_year_df['size']\n",
    "    y_mask = y_mask[crime_type_year_df['Primary Type'] == mask]\n",
    "    \n",
    "    crime_string = str(mask)\n",
    "    \n",
    "    # make label lowercase\n",
    "    crime_string = crime_string.lower()\n",
    "    \n",
    "    # capitalize first letter \n",
    "    crime_string = crime_string.capitalize()\n",
    "\n",
    "    mySelectedLabel.value =  crime_string + ' ' +  'crimes reported:' \\\n",
    "    + ' '+ str(sum(y_mask))\n",
    "                                            \n",
    "    line_plot.x = x_mask\n",
    "    line_plot.y = y_mask\n",
    "\n",
    "bar.observe(get_tot_crime_types, 'selected')\n",
    "\n",
    "ipywidgets.VBox([mySelectedLabel,ipywidgets.HBox([fig, fig2])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Deceptive Practice: This crime has a very wide field of coverage, but a simple definition is that this crime is commited when one attempts to defraud another person. \n",
    "\n",
    "All of the figures in this report use data from Crimes - 2001 to present datset, which contains all recorded and reported crime data in Chicago from 2001 to 2019. This includes location, time, type of incident, description, arrest True/False boolean, and so on. However, my visualizations are only concerned with the year, crime type, and the arrest status of the incident. The first visualization is an interactive dashboard. If you click on a crime type bar, you should be shown how many incidents of that type have occurred from 2001 to 2019 in Chicago, but on the right hand side the line plot will also update where it will show how many incidents there were for that crime per year. When it comes to come in Chicago, the talking point aims upon the violent crime on Chicago, but this visualization provides more context to the situation in which homicide does not even make the ten most common crime type list. In fact,homocide is the 21th most common crime in Chicago, and this list is dominated by non-violent crimes. Lastly, it is plain to see by looking at the line plot that there is a current downward trend in Chicago crime frequency. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # remove domestic violence because there  is only one  incident in the entire file\n",
    "# arrest_df = crimes_df.groupby(['Primary Type', 'Arrest']).size().reset_index()\n",
    "# # arrest_df = arrest_df[arrest_df['Primary Type'] != 'DOMESTIC VIOLENCE']\n",
    "\n",
    "# arrest_df\n",
    "# arrest_df_F = arrest_df[arrest_df['Arrest'] == False]\\\n",
    "# .sort_values(by=0,ascending=False).rename(columns={0: 'false_size'})\n",
    "# arrest_df_T = arrest_df[arrest_df['Arrest'] == True]\\\n",
    "# .sort_values(by=0,ascending=False).rename(columns={0: 'true_size'})\n",
    "\n",
    "# arrest_perc_df = pd.merge(arrest_df_T, arrest_df_F, left_on='Primary Type', right_on='Primary Type', how='inner')\n",
    "# arrest_perc_df['Arrest Made %'] = arrest_perc_df['true_size'] / (arrest_perc_df['false_size'] + arrest_perc_df['true_size'] )\n",
    "# arrest_perc_df = arrest_perc_df.sort_values(by='Arrest Made %', ascending=False).reset_index(drop=True)\n",
    "# arrest_perc_df = arrest_perc_df[['Primary Type', 'Arrest Made %']]\n",
    "\n",
    "# arrest_perc_df.to_csv('arrest_perc_df.csv')\n",
    "# arrest_perc_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Primary Type</th>\n",
       "      <th>Arrest Made %</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>PROSTITUTION</td>\n",
       "      <td>0.996105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>NARCOTICS</td>\n",
       "      <td>0.994111</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>PUBLIC INDECENCY</td>\n",
       "      <td>0.993902</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>GAMBLING</td>\n",
       "      <td>0.992795</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>LIQUOR LAW VIO</td>\n",
       "      <td>0.991148</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       Primary Type  Arrest Made %\n",
       "0      PROSTITUTION       0.996105\n",
       "1         NARCOTICS       0.994111\n",
       "2  PUBLIC INDECENCY       0.993902\n",
       "3          GAMBLING       0.992795\n",
       "4    LIQUOR LAW VIO       0.991148"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# read in aggregate csv file\n",
    "arrest_perc_df = pd.read_csv(\"arrest_perc_df.csv\").drop(columns='Unnamed: 0')\n",
    "arrest_perc_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b2884e3298f84c1db645ffec0b8e9a41",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "VBox(children=(Label(value=''), HBox(children=(Figure(axes=[Axis(label='Type of Crime', offset={'scale': Ordin…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "x2 = arrest_perc_df['Primary Type']\n",
    "y2 = arrest_perc_df['Arrest Made %']\n",
    "\n",
    "x_sc3 = bqplot.OrdinalScale()\n",
    "y_sc3 = bqplot.LinearScale()\n",
    "\n",
    "\n",
    "bar2 = bqplot.Bars(x=x2, y = y2, # set y-values to per thousand, so values are more clear.\n",
    "                scales={'x': x_sc3, 'y': y_sc3}, \n",
    "                interactions={'click':'select'},\n",
    "                selected_style={'fill': 'orange'})\n",
    "\n",
    "ax_x3 = bqplot.Axis(scale=x_sc3, \n",
    "                   tick_style= {'font-size': '8px'}, # set font-size to seven for increased legibility\n",
    "                   tick_rotate = -700,  # rotate x-axis values\n",
    "                   offset={'scale': x_sc3, 'value': '50'}, # offset x-axis values\n",
    "                   label= 'Type of Crime')  # set x-axis label\n",
    "\n",
    "ax_y3 = bqplot.Axis(scale=y_sc3,\n",
    "                   orientation='vertical', \n",
    "                   label='Arrest Percentage', # set y-axis label\n",
    "                   label_offset='-50') # y-axis label offset\n",
    "\n",
    "# initiate a label above the plot\n",
    "mySelectedLabel2 = ipywidgets.Label()\n",
    "\n",
    "# initiate plot year label range\n",
    "year_range2 = bqplot.Label(x=[0.75],y=[0.10], font_size=52, font_weight='bolder', \n",
    "                         colors=['blue'], \n",
    "                          text=[('Year Range: 2001 to 2019')], \n",
    "                          x_offset = -125,\n",
    "                          y_offset = -300)\n",
    "    \n",
    "fig3 = bqplot.Figure(marks=[bar2, year_range2], axes=[ax_x3, ax_y3],\n",
    "                    title=\"Chicagoan Reported Crimes by Arrest Percentage\")\n",
    "def get_crime_arrest_perc(change2):\n",
    "    x_values = x2[change2['owner'].selected[0]]\n",
    "    \n",
    "    y_values = y2[change2['owner'].selected[0]]\n",
    "    y_values = y_values.round(3)\n",
    "    y_values = str(y_values * 100)\n",
    "  \n",
    "    mySelectedLabel2.value =  str(x_values.lower().capitalize()) + ' ' + \\\n",
    "    'reported crimes arrest percentage: ' + y_values + '%'\n",
    "                            \n",
    "fig3.layout.width = '900px'\n",
    "\n",
    "bar2.observe(get_crime_arrest_perc, 'selected')\n",
    "ipywidgets.VBox([mySelectedLabel2,ipywidgets.HBox([fig3])])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Deceptive Practice: This crime has a very wide field of coverage, but a simple definition is that this crime is commited when one attempts to defraud another person.\n",
    "\n",
    "Also, take note that VIO stands in place of violation.\n",
    "\n",
    "While the dashboard provides more context into which crimes are the most frequent in Chicago, this visualization offers a differnet kind of context. It illustrates that some reported crimes have very little difficult in making arrests whereas others, such as theft, are very difficult. One might argue that theft has such a low arrest percentage report due to so many thefts, but narcotics is the 4th most common crime, and it is also the crime that has the second highest arrrest percentage at almost 100%. Hence, this visualization illustrates that crime frequency does not have a direct relationship with arrest rate per crime. Interestingly enough it is the crimes that are often considered misdemeanors that have the lowest crime arrest percentage, which makes sense becuase they are not often high priorities. In conclusion, this visualization gives us a better understanding of why certain crimes, such as homicide, get so much attention because it is difficult to make arrests for these crimes. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "# la_crime_df = pd.read_csv(\"Crime_Data_from_2010_to_Present.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# la_crime_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # extract year from date occured\n",
    "# la_crime_df['Date Occurred'] = la_crime_df['Date Occurred'].str[6:]\n",
    "# la_crime_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # make la_crime_df a string then int \n",
    "# la_crime_df['Date Occurred'] = la_crime_df['Date Occurred'].astype(str)\n",
    "# la_crime_df['Date Occurred'] = la_crime_df['Date Occurred'].astype(int)\n",
    "\n",
    "# # subset  date occured as less than 2019 because a full year has not elapsed\n",
    "# la_crime_df = la_crime_df[la_crime_df['Date Occurred'] < 2019]\n",
    "\n",
    "# la_crime_type_size_df = la_crime_df.groupby('Date Occurred').size()\n",
    "# la_crime_type_size_df.to_csv('la_crime_type_size_df.csv')\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0\n",
       "2010    208787\n",
       "2011    200407\n",
       "2012    201124\n",
       "2013    192163\n",
       "2014    194956\n",
       "Name: 1, dtype: int64"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# I had trouble reading in the series from a csv, and I used code from the following web page:\n",
    "# https://stackoverflow.com/questions/42597368/when-i-read-in-a-pandas-series-from-csv-it-is-turning-it-into-a-dataframe\n",
    "\n",
    "# read in aggregate csv file\n",
    "la_crime_type_size_df = pd.read_csv(\"la_crime_type_size_df.csv\", header=None, squeeze=True, index_col=0)  \n",
    "la_crime_type_size_df.head()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # subset year from 2010 to 2018 to match LA comparison--exlcude 2019 because a full year has not elapsed\n",
    "# crimes_2010_2018_df = crimes_df[(crimes_df['Year'] >= 2010) & (crimes_df['Year'] < 2019)]\n",
    "\n",
    "# crimes_2010_2018_df = crimes_2010_2018_df.groupby('Year').size()\n",
    "# crimes_2010_2018_df.to_csv('crimes_2010_2018_df.csv')\n",
    "# crimes_2010_2018_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0\n",
       "2010    370313\n",
       "2011    351777\n",
       "2012    335970\n",
       "2013    307100\n",
       "2014    275289\n",
       "Name: 1, dtype: int64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# I had trouble reading in the series from a csv, and I used code from the following web page:\n",
    "# https://stackoverflow.com/questions/42597368/when-i-read-in-a-pandas-series-from-csv-it-is-turning-it-into-a-dataframe\n",
    "\n",
    "# read in aggregate csv file\n",
    "crimes_2010_2018_df = pd.read_csv(\"crimes_2010_2018_df.csv\", header=None, squeeze=True, index_col=0)  \n",
    "\n",
    "crimes_2010_2018_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "with plt.style.context(\"ggplot\"):\n",
    "    plt.plot(crimes_2010_2018_df, marker='.', linewidth = 1.0, label=\"Chicago\")\n",
    "    plt.plot(la_crime_type_size_df, marker='.', linewidth = 1.0, label=\"Los Angeles\")\n",
    "    \n",
    "    plt.fill_between(la_crime_type_size_df.index, crimes_2010_2018_df, la_crime_type_size_df, color = \"#aaaaaa\")\n",
    "\n",
    "    plt.title(\"Chicago vs. Los Angeles Reported Crimes by Year\", fontsize=12)\n",
    "    \n",
    "    x_index = np.arange(2010, 2019)\n",
    "    y_index = np.arange(175000, 400000, 50000)\n",
    "    plt.xticks(x_index)\n",
    "    plt.yticks(y_index)\n",
    "    plt.xlabel(\"Year\")\n",
    "    plt.ylabel(\"Total Reported Crimes\")\n",
    "    \n",
    "    plt.legend()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "While the first two visualizations provides context for which types of crime are the most frequent in Chicago and why some crimes might be more polarized in the media, this visualization clearly illustrates the trend that was hinted to in the dashboard visualization where Chicagoan crime has severely dropped in recent years. Within an eight year time span there has been a decrease of about 100,000 crimes per year. Furthermore, the inclusion of Los Angeles in the last visualization, which is city of slightly larger magnitude in population, demonstrates that Chicago is not the only large metropolitan city in the United States faced with significant levels of crime. The intention of these visualizations is not to say that Chicagoan crime is not a serious issue, it is. However, these visualizations illustrate that the overall severity of Chicagoan crime is overstated and that Chicago is not alone in the United States when it comes to defeating with crime. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Bibliography\n",
    "\n",
    "Police. Crimes - 2001 to present. April 17, 2001. Distributed by Chicago Data Portal. https://data.cityofchicago.org/Public-Safety/Crimes-2001-to-present/ijzp-q8t2. \n",
    "\n",
    "LAPD. Crime Data from 2010 to Present. April 10, 2017. Distributed by Los Angeles Open Data. https://data.lacity.org/A-Safe-City/Crime-Data-from-2010-to-Present/y8tr-7khq. \n",
    "\n",
    "\n"
   ]
  }
 ],
 "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": "3.7.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}