analysis.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Women homicide analysis\n",
    "\n",
    "By [Iris Lee](http://www.twitter.com/@irisslee)  \n",
    "  \n",
    "The Los Angeles Times [Homicide Report](https://homicide.latimes.com) conducted an analysis of female victims. It found: \n",
    "\n",
    "* The number of women being killed is increasing.\n",
    "* Of the 295 women killed over past the three years, 75% of them are \"cleared\" - meaning suspect has been identified - by an investigating agency.  \n",
    "* Of the 214 cleared cases, 170 cases had a suspect that knew the victim prior to killing.\n",
    "* There were 99 cases where the suspect was in an intimate relationship with the victim.\n",
    "\n",
    "These findings resulted in the September 19th story [\"As homicides drop in L.A., more women are being killed — often by intimate partners\"](https://www.latimes.com/projects/women-violence-homicides-increase-death-murder/)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Configuration  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import warnings\n",
    "import pandas as pd\n",
    "import altair as alt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DataTransformerRegistry.enable('json')"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "warnings.filterwarnings('ignore')\n",
    "pd.set_option('display.max_columns',500)\n",
    "alt.data_transformers.enable('json')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Download and read homicides  \n",
    "Retrieves homicide list from the Times' unpublished API. This step had been withheld from the notebook to maintain our site's security."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "homicides = pd.read_csv(os.getenv(\"HOMICIDE_REPORT_API\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate out death year from death date"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "homicides[\"death date\"] = pd.to_datetime(homicides['death date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "homicides['death_year'] = homicides['death date'].dt.year"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate out 10 years"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "h10 = homicides[homicides[\"death_year\"].isin(list(range(2008,2019)))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ID</th>\n",
       "      <th>coroners case number</th>\n",
       "      <th>police case number</th>\n",
       "      <th>da case number</th>\n",
       "      <th>slug</th>\n",
       "      <th>first name</th>\n",
       "      <th>middle name</th>\n",
       "      <th>last name</th>\n",
       "      <th>suffix</th>\n",
       "      <th>death date</th>\n",
       "      <th>death time</th>\n",
       "      <th>age</th>\n",
       "      <th>date of birth</th>\n",
       "      <th>cause</th>\n",
       "      <th>security hold</th>\n",
       "      <th>Have autopsy?</th>\n",
       "      <th>gender</th>\n",
       "      <th>race</th>\n",
       "      <th>died on scene</th>\n",
       "      <th>sequela</th>\n",
       "      <th>incident address</th>\n",
       "      <th>incident zipcode</th>\n",
       "      <th>officer involved</th>\n",
       "      <th>ruling</th>\n",
       "      <th>coroners description</th>\n",
       "      <th>synopsis</th>\n",
       "      <th>incident point x</th>\n",
       "      <th>incident point y</th>\n",
       "      <th>neighborhoodv6__name</th>\n",
       "      <th>jurisdiction__name</th>\n",
       "      <th>image</th>\n",
       "      <th>death_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>365</th>\n",
       "      <td>16949</td>\n",
       "      <td>2018-09898</td>\n",
       "      <td>018-19377-0459-011</td>\n",
       "      <td>NaN</td>\n",
       "      <td>jose-carlos-soto-jr</td>\n",
       "      <td>Jose</td>\n",
       "      <td>Carlos</td>\n",
       "      <td>Soto</td>\n",
       "      <td>Jr.</td>\n",
       "      <td>2018-12-31</td>\n",
       "      <td>17:06:00</td>\n",
       "      <td>26.0</td>\n",
       "      <td>1992-03-03</td>\n",
       "      <td>gunshot</td>\n",
       "      <td>False</td>\n",
       "      <td>False</td>\n",
       "      <td>male</td>\n",
       "      <td>latino</td>\n",
       "      <td>True</td>\n",
       "      <td>False</td>\n",
       "      <td>14300 block of Carmenita Road</td>\n",
       "      <td>90650.0</td>\n",
       "      <td>False</td>\n",
       "      <td>H</td>\n",
       "      <td>GUNSHOT WOUNDS OF THE CHEST    26Y/O MALE WAS ...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>-118.046626</td>\n",
       "      <td>33.901736</td>\n",
       "      <td>Norwalk</td>\n",
       "      <td>LASD Homicide</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2018</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "        ID coroners case number  police case number da case number  \\\n",
       "365  16949           2018-09898  018-19377-0459-011            NaN   \n",
       "\n",
       "                    slug first name middle name last name suffix death date  \\\n",
       "365  jose-carlos-soto-jr       Jose      Carlos      Soto    Jr. 2018-12-31   \n",
       "\n",
       "    death time   age date of birth    cause  security hold  Have autopsy?  \\\n",
       "365   17:06:00  26.0    1992-03-03  gunshot          False          False   \n",
       "\n",
       "    gender    race  died on scene sequela               incident address  \\\n",
       "365   male  latino           True   False  14300 block of Carmenita Road   \n",
       "\n",
       "     incident zipcode  officer involved ruling  \\\n",
       "365           90650.0             False      H   \n",
       "\n",
       "                                  coroners description synopsis  \\\n",
       "365  GUNSHOT WOUNDS OF THE CHEST    26Y/O MALE WAS ...      NaN   \n",
       "\n",
       "     incident point x  incident point y neighborhoodv6__name  \\\n",
       "365       -118.046626         33.901736              Norwalk   \n",
       "\n",
       "    jurisdiction__name image  death_year  \n",
       "365      LASD Homicide   NaN        2018  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h10.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "7505"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h10.groupby('death_year').size().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate out victims with gender listed "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "with_gender = h10[h10.gender.isin(['male','female'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "male      6395\n",
       "female    1096\n",
       "Name: gender, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with_gender.gender.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "male      0.853691\n",
       "female    0.146309\n",
       "Name: gender, dtype: float64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with_gender.gender.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>gender</th>\n",
       "      <th>female</th>\n",
       "      <th>male</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>death_year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2008</th>\n",
       "      <td>131</td>\n",
       "      <td>755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2009</th>\n",
       "      <td>108</td>\n",
       "      <td>660</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2010</th>\n",
       "      <td>121</td>\n",
       "      <td>568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2011</th>\n",
       "      <td>96</td>\n",
       "      <td>548</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2012</th>\n",
       "      <td>89</td>\n",
       "      <td>567</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2013</th>\n",
       "      <td>76</td>\n",
       "      <td>546</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2014</th>\n",
       "      <td>82</td>\n",
       "      <td>503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2015</th>\n",
       "      <td>98</td>\n",
       "      <td>566</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2016</th>\n",
       "      <td>87</td>\n",
       "      <td>609</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2017</th>\n",
       "      <td>94</td>\n",
       "      <td>550</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2018</th>\n",
       "      <td>114</td>\n",
       "      <td>523</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "gender      female  male\n",
       "death_year              \n",
       "2008           131   755\n",
       "2009           108   660\n",
       "2010           121   568\n",
       "2011            96   548\n",
       "2012            89   567\n",
       "2013            76   546\n",
       "2014            82   503\n",
       "2015            98   566\n",
       "2016            87   609\n",
       "2017            94   550\n",
       "2018           114   523"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "with_gender.groupby(['death_year', 'gender']).size().unstack(level=-1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How has number of homicide victims change over the years?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.vegalite.v3+json": {
       "$schema": "https://vega.github.io/schema/vega-lite/v3.4.0.json",
       "config": {
        "mark": {
         "tooltip": null
        },
        "view": {
         "height": 300,
         "width": 400
        }
       },
       "data": {
        "format": {
         "type": "json"
        },
        "url": "altair-data-66ab408643827f31e2e3cc809488153d.json"
       },
       "layer": [
        {
         "encoding": {
          "color": {
           "field": "gender",
           "sort": "descending",
           "type": "nominal"
          },
          "order": {
           "field": "gender",
           "sort": "ascending",
           "type": "nominal"
          },
          "x": {
           "axis": {
            "title": "Year"
           },
           "field": "death_year",
           "type": "ordinal"
          },
          "y": {
           "aggregate": "count",
           "axis": {
            "title": "Homicide count"
           },
           "field": "ID",
           "type": "quantitative"
          }
         },
         "mark": "bar",
         "title": "Homicide victims: 2008-2018",
         "width": 500
        },
        {
         "encoding": {
          "color": {
           "field": "gender",
           "sort": "descending",
           "type": "nominal"
          },
          "order": {
           "field": "gender",
           "sort": "ascending",
           "type": "nominal"
          },
          "text": {
           "aggregate": "count",
           "field": "ID",
           "type": "quantitative"
          },
          "x": {
           "axis": {
            "title": "Year"
           },
           "field": "death_year",
           "type": "ordinal"
          },
          "y": {
           "aggregate": "count",
           "axis": {
            "title": "Homicide count"
           },
           "field": "ID",
           "type": "quantitative"
          }
         },
         "mark": {
          "align": "center",
          "dy": -5,
          "stroke": "black",
          "strokeWidth": 0.5,
          "type": "text"
         },
         "title": "Homicide victims: 2008-2018",
         "width": 500
        }
       ]
      },
      "image/png": "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",
      "text/plain": [
       "<VegaLite 3 object>\n",
       "\n",
       "If you see this message, it means the renderer has not been properly enabled\n",
       "for the frontend that you are using. For more information, see\n",
       "https://altair-viz.github.io/user_guide/troubleshooting.html\n"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ten = alt.Chart(with_gender, title=\"Homicide victims: 2008-2018\").mark_bar().encode(\n",
    "        x=alt.X('death_year:O', axis = alt.Axis(title = \"Year\")),\n",
    "        y = alt.Y('count(ID)', axis = alt.Axis(title = \"Homicide count\")),\n",
    "        color= alt.Color('gender:N',sort=alt.SortOrder('descending')),\n",
    "        order = alt.Order('gender', sort='ascending')\n",
    "    ).properties(width = 500)\n",
    "\n",
    "text = ten.mark_text(\n",
    "        align = 'center',\n",
    "        stroke = 'black',\n",
    "        strokeWidth = 0.5,\n",
    "        dy= -5\n",
    "    ).encode(text ='count(ID)')\n",
    "\n",
    "ten + text"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **While fewer and fewer young men are being killed on the streets of LA, the number of female victims has increased.** "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Women homicides only from 2016-2018"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "_1618 = with_gender[(with_gender.gender == 'female') & (with_gender.death_year.isin([2016,2017,2018]))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "295"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "_1618.ID.nunique()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **In the past three years, 295 women were killed across the county**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Review the police investigation of all 295 female victims"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Times then reviewed the outcome of the police investigation for all female victims from 2016 to 2018.\n",
    "\n",
    "We confirmed the following through records requests and interviews with the investigating agencies. \n",
    "1. Is the case being investigated by law enforcement?  \n",
    "2. What is the status of the investigation?  \n",
    "3. For \"cleared\" cases - the ones where the suspect is known to law enforcement - what was the suspect-victim relationship?\n",
    "\n",
    "The data is not be published with this notebook to avoid airing sensitive and potentially inaccurate information that was not fit for inclusion in our story."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "victims =  pd.read_csv('victims.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate out the death year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "victims[\"death date\"] = pd.to_datetime(victims['death date'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "victims['death_year'] = victims[\"death date\"].dt.year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "295"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(victims)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Verify our row count matches the homicide report database"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert len(victims) == len(_1618) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "assert victims['coroners case number'].isin(_1618['coroners case number']).all()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "From the victims reviewed, we filtered out the ones law enforcement did not pursue (accidents, sequela..etc ).\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# c = cleared\n",
    "# U = open investigation\n",
    "cases = victims[victims.cleared.isin(['c','U'])]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "291"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cases)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **The police investigated 291 cases from 2016-2018**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "c    214\n",
       "U     77\n",
       "Name: cleared, dtype: int64"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.cleared.value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "c    0.735395\n",
       "U    0.264605\n",
       "Name: cleared, dtype: float64"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.cleared.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Filter out only cleared cases where law enforcement have a suspect in mind."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "cleared = cases[cases.cleared == 'c']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7353951890034365"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cleared)/len(cases)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **Nearly 3 out of 4 cases, the police have names a suspect (74%)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Y    0.794393\n",
       "N    0.205607\n",
       "Name: suspect_knew_victim, dtype: float64"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleared['suspect_knew_victim'].value_counts('normalize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Y    170\n",
       "N     44\n",
       "Name: suspect_knew_victim, dtype: int64"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleared['suspect_knew_victim'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **The review found that in cases where police have named the suspect, the vast majority of accused killers knew their victim.**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Separate out intimate relationships. List all determined "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "husband                   46\n",
       "boyfriend                 23\n",
       "ex-boyfriend              21\n",
       "son                       11\n",
       "OIS                        8\n",
       "acquaintance               7\n",
       "friend                     6\n",
       "partner                    6\n",
       "father                     5\n",
       "mother                     4\n",
       "roomate                    4\n",
       "mother's ex-boyfriend      3\n",
       "ex-husband                 3\n",
       "caregiver                  3\n",
       "granddaughter              2\n",
       "nephew                     2\n",
       "client                     2\n",
       "brother                    2\n",
       "great granddaughter        1\n",
       "gang                       1\n",
       "parents' co-worker         1\n",
       "daughter's boyfriend       1\n",
       "son's friend               1\n",
       "friend's ex-husband        1\n",
       "friend's brother           1\n",
       "stepsister's boyfriend     1\n",
       "son-in-law                 1\n",
       "sister's ex-boyfriend      1\n",
       "renter on property         1\n",
       "tenant's niece             1\n",
       "mother's boyfriend         1\n",
       "grandson                   1\n",
       "daughter's ex              1\n",
       "sister's boyfriend         1\n",
       "caregiver's son            1\n",
       "neighbor                   1\n",
       "grandmother                1\n",
       "brother-in-law             1\n",
       "Name: suspect_relationship, dtype: int64"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cleared[\"suspect_relationship\"].value_counts()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "intimate_relationships = ['husband', 'boyfriend', 'ex-boyfriend', 'ex-husband','partner']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "intimate_homicides = cleared[cleared['suspect_relationship'].isin(intimate_relationships)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "99"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(intimate_homicides)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.46261682242990654"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(intimate_homicides)/len(cleared)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **Nearly half were, at some point, in an intimate relationship; husbands, boyfriends and, often, exes.(46%)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "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>coroners case number</th>\n",
       "      <th>first name</th>\n",
       "      <th>middle name</th>\n",
       "      <th>last name</th>\n",
       "      <th>death date</th>\n",
       "      <th>age</th>\n",
       "      <th>cause</th>\n",
       "      <th>race</th>\n",
       "      <th>notes</th>\n",
       "      <th>cleared</th>\n",
       "      <th>suspect_knew_victim</th>\n",
       "      <th>suspect_relationship</th>\n",
       "      <th>coverage</th>\n",
       "      <th>death_year</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>2017-02882</td>\n",
       "      <td>Alexiz</td>\n",
       "      <td>Miranda</td>\n",
       "      <td>Orona</td>\n",
       "      <td>2017-04-14</td>\n",
       "      <td>18</td>\n",
       "      <td>gunshot</td>\n",
       "      <td>latino</td>\n",
       "      <td>boyfriend +3 more suspects</td>\n",
       "      <td>c</td>\n",
       "      <td>Y</td>\n",
       "      <td>boyfriend</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2017.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  coroners case number first name middle name last name death date  age  \\\n",
       "3           2017-02882     Alexiz     Miranda     Orona 2017-04-14   18   \n",
       "\n",
       "     cause    race                       notes cleared suspect_knew_victim  \\\n",
       "3  gunshot  latino  boyfriend +3 more suspects       c                   Y   \n",
       "\n",
       "  suspect_relationship coverage  death_year  \n",
       "3            boyfriend      NaN      2017.0  "
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intimate_homicides.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2017.0    40\n",
       "2018.0    34\n",
       "2016.0    25\n",
       "Name: death_year, dtype: int64"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "intimate_homicides.death_year.value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Create column to mark intimate partners"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "cases['ip'] = cases['suspect_relationship'].isin(intimate_relationships)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "death_year  suspect_knew_victim\n",
       "2016.0      Y                      0.738462\n",
       "            N                      0.261538\n",
       "2017.0      Y                      0.842857\n",
       "            N                      0.157143\n",
       "2018.0      Y                      0.807692\n",
       "            N                      0.192308\n",
       "Name: suspect_knew_victim, dtype: float64"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.groupby('death_year')['suspect_knew_victim'].value_counts('normalize')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**CAUSES**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gunshot        0.591696\n",
       "stabbing       0.169550\n",
       "blunt-force    0.124567\n",
       "other          0.055363\n",
       "strangled      0.038062\n",
       "pending        0.020761\n",
       "Name: cause, dtype: float64"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.cause.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gunshot        0.632653\n",
       "stabbing       0.214286\n",
       "blunt-force    0.081633\n",
       "strangled      0.040816\n",
       "other          0.030612\n",
       "Name: cause, dtype: float64"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases[cases.ip == True].cause.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**RACE**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "latino    0.440972\n",
       "black     0.250000\n",
       "white     0.225694\n",
       "asian     0.079861\n",
       "other     0.003472\n",
       "Name: race, dtype: float64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.race.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "latino    0.443299\n",
       "white     0.309278\n",
       "black     0.164948\n",
       "asian     0.082474\n",
       "Name: race, dtype: float64"
      ]
     },
     "execution_count": 44,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases[cases.ip == True].race.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**AGE**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35.0"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.age.median()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "35.0"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases[cases.ip == True].age.median()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Causes and age among male victims vs. female victims from 2016 to 2018"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "men_1618 = with_gender[(with_gender.gender == 'male') & (with_gender.death_year.isin([2016,2017,2018]))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "gunshot         0.769185\n",
       "stabbing        0.104105\n",
       "blunt-force     0.078525\n",
       "other           0.033314\n",
       "pending         0.007139\n",
       "strangled       0.005354\n",
       "undetermined    0.002380\n",
       "Name: cause, dtype: float64"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "men_1618.cause.value_counts('normalize')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "31.0"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "men_1618.age.median()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Income levels in incident neighborhoods "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Merge case information with previous API pull for complete information"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "cases_detailed = pd.merge(cases, _1618, on='coroners case number', how='inner')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "291"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cases_detailed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "What are top neighborhoods where homicide occurs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['Long Beach', 'Inglewood', 'Westlake', 'Compton', 'Carson', 'Hawthorne']"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases_detailed['neighborhoodv6__name'].value_counts().head(6).index.tolist()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Read median income file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "income = pd.read_csv('2017_income.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "270"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(income)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Median income for L.A. County is $61,015 per [U.S. Census]('https://www.census.gov/quickfacts/losangelescountycalifornia')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Neighborhoods below median income level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "below = income[income.median_income < 61015].neighborhood.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "113"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(below)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Neighborhoods above median income level"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "above = income[income.median_income > 61015].neighborhood.tolist()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "157"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(above)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "how many victims' neighborhood is from ones below median?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "190"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cases_detailed[cases_detailed['neighborhoodv6__name'].isin(below)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "how many victims' neighborhood from ones above median income"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "89"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cases_detailed[cases_detailed['neighborhoodv6__name'].isin(above)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "How many with missing neighborhoods?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "12"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(cases_detailed)-(len(cases_detailed[cases_detailed['neighborhoodv6__name'].isin(below)])+len(cases_detailed[cases_detailed['neighborhoodv6__name'].isin(above)]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6810035842293907"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(190)/(291-12)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "=> **70% victims died in neighborhoods with median income lower than county's median**"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Compare male victims"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.7829964328180737"
      ]
     },
     "execution_count": 70,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(men_1618[men_1618['neighborhoodv6__name'].isin(below)])/len(men_1618)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create a dataset ready for visualization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Modify dataframe for visualization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "viz = cases.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "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>coroners case number</th>\n",
       "      <th>first name</th>\n",
       "      <th>middle name</th>\n",
       "      <th>last name</th>\n",
       "      <th>death date</th>\n",
       "      <th>age</th>\n",
       "      <th>cause</th>\n",
       "      <th>race</th>\n",
       "      <th>notes</th>\n",
       "      <th>cleared</th>\n",
       "      <th>suspect_knew_victim</th>\n",
       "      <th>suspect_relationship</th>\n",
       "      <th>coverage</th>\n",
       "      <th>death_year</th>\n",
       "      <th>ip</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>2018-05087</td>\n",
       "      <td>Abril</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Onofre</td>\n",
       "      <td>2018-07-05</td>\n",
       "      <td>19</td>\n",
       "      <td>gunshot</td>\n",
       "      <td>latino</td>\n",
       "      <td>NaN</td>\n",
       "      <td>U</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>https://homicide.latimes.com/post/abril-onofre/</td>\n",
       "      <td>2018.0</td>\n",
       "      <td>False</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  coroners case number first name middle name last name death date  age  \\\n",
       "0           2018-05087      Abril         NaN    Onofre 2018-07-05   19   \n",
       "\n",
       "     cause    race notes cleared suspect_knew_victim suspect_relationship  \\\n",
       "0  gunshot  latino   NaN       U                 NaN                  NaN   \n",
       "\n",
       "                                          coverage  death_year     ip  \n",
       "0  https://homicide.latimes.com/post/abril-onofre/      2018.0  False  "
      ]
     },
     "execution_count": 72,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cases.head(1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Mark open investigation\n",
    "viz.loc[viz.cleared=='U', 'status'] = 'open'\n",
    "#Mark ones with arrest but suspect is random to victim\n",
    "viz.loc[((viz.cleared=='c')& ~(viz['suspect_knew_victim'] == 'Y')), 'status'] = 'random_suspect'\n",
    "#Mark ones with arrest and suspect is known to victim \n",
    "viz.loc[((viz.cleared=='c')&(viz['suspect_knew_victim'] == 'Y')), 'status'] = 'familiar_suspect'\n",
    "#mark ones with arresta and suspect is romantic\n",
    "viz.loc[((viz.cleared=='c')&(viz['ip'] == True)), 'status'] = 'intimate_partner'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmed_viz = viz.rename(columns={'last name':'last_name','death date':'death_date' })"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmed_viz = trimmed_viz[[\n",
    "    'last_name',\n",
    "    'death_date',\n",
    "    'age',\n",
    "    'cause',\n",
    "    'race',\n",
    "    'status'\n",
    "]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "trimmed_viz.to_csv(\"women.csv\", index=False)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}