{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:14.848930Z",
"start_time": "2018-10-02T04:48:13.640150Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
" \n",
"\n",
"\n",
" \n",
"\n",
"
\n",
"
\n",
"
BokehJS successfully loaded.\n",
"
"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"# Custom libraries\n",
"from datascienceutils import plotter\n",
"from datascienceutils import analyze\n",
"\n",
"# Standard libraries\n",
"import json\n",
"%matplotlib inline\n",
"import datetime\n",
"import numpy as np\n",
"import pandas as pd\n",
"import random\n",
"\n",
"\n",
"from bokeh.plotting import figure, show, output_file, output_notebook, ColumnDataSource\n",
"from bokeh.charts import Histogram\n",
"import bokeh\n",
"output_notebook()\n"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:14.858373Z",
"start_time": "2018-10-02T04:48:14.850657Z"
},
"scrolled": false
},
"outputs": [],
"source": [
"murder_2015 = pd.read_csv('./data/fivethirtyeight_data/murder_2016/murder_2015_final.csv')\n",
"murder_2016 = pd.read_csv('./data/fivethirtyeight_data/murder_2016/murder_2016_prelim.csv')\n"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:14.983752Z",
"start_time": "2018-10-02T04:48:14.859819Z"
},
"scrolled": false
},
"outputs": [],
"source": [
"numericalCols = murder_2015.select_dtypes(include=[np.number]).columns\n",
"catCols = set(murder_2015.columns) -set(numericalCols)"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"ExecuteTime": {
"end_time": "2018-09-29T10:16:18.290573Z",
"start_time": "2018-09-29T10:16:18.284475Z"
},
"scrolled": false
},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.084677Z",
"start_time": "2018-10-02T04:48:14.985085Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" 2014_murders | \n",
" 2015_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" count | \n",
" 83.000000 | \n",
" 83.000000 | \n",
" 83.000000 | \n",
"
\n",
" \n",
" mean | \n",
" 65.746988 | \n",
" 75.481928 | \n",
" 9.734940 | \n",
"
\n",
" \n",
" std | \n",
" 79.011244 | \n",
" 91.684289 | \n",
" 21.858795 | \n",
"
\n",
" \n",
" min | \n",
" 0.000000 | \n",
" 1.000000 | \n",
" -19.000000 | \n",
"
\n",
" \n",
" 25% | \n",
" 19.500000 | \n",
" 22.500000 | \n",
" -3.000000 | \n",
"
\n",
" \n",
" 50% | \n",
" 32.000000 | \n",
" 39.000000 | \n",
" 4.000000 | \n",
"
\n",
" \n",
" 75% | \n",
" 82.000000 | \n",
" 94.000000 | \n",
" 14.000000 | \n",
"
\n",
" \n",
" max | \n",
" 411.000000 | \n",
" 478.000000 | \n",
" 133.000000 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" 2014_murders 2015_murders change\n",
"count 83.000000 83.000000 83.000000\n",
"mean 65.746988 75.481928 9.734940\n",
"std 79.011244 91.684289 21.858795\n",
"min 0.000000 1.000000 -19.000000\n",
"25% 19.500000 22.500000 -3.000000\n",
"50% 32.000000 39.000000 4.000000\n",
"75% 82.000000 94.000000 14.000000\n",
"max 411.000000 478.000000 133.000000"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2015.describe()"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.175389Z",
"start_time": "2018-10-02T04:48:15.085965Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" 2015_murders | \n",
" 2016_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" count | \n",
" 79.000000 | \n",
" 79.000000 | \n",
" 79.000000 | \n",
"
\n",
" \n",
" mean | \n",
" 56.468354 | \n",
" 62.379747 | \n",
" 5.911392 | \n",
"
\n",
" \n",
" std | \n",
" 70.158764 | \n",
" 81.051472 | \n",
" 21.962199 | \n",
"
\n",
" \n",
" min | \n",
" 0.000000 | \n",
" 1.000000 | \n",
" -21.000000 | \n",
"
\n",
" \n",
" 25% | \n",
" 13.000000 | \n",
" 14.500000 | \n",
" -3.000000 | \n",
"
\n",
" \n",
" 50% | \n",
" 30.000000 | \n",
" 30.000000 | \n",
" 2.000000 | \n",
"
\n",
" \n",
" 75% | \n",
" 71.500000 | \n",
" 81.500000 | \n",
" 9.000000 | \n",
"
\n",
" \n",
" max | \n",
" 378.000000 | \n",
" 536.000000 | \n",
" 158.000000 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" 2015_murders 2016_murders change\n",
"count 79.000000 79.000000 79.000000\n",
"mean 56.468354 62.379747 5.911392\n",
"std 70.158764 81.051472 21.962199\n",
"min 0.000000 1.000000 -21.000000\n",
"25% 13.000000 14.500000 -3.000000\n",
"50% 30.000000 30.000000 2.000000\n",
"75% 71.500000 81.500000 9.000000\n",
"max 378.000000 536.000000 158.000000"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2016.describe()"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.259874Z",
"start_time": "2018-10-02T04:48:15.176800Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" city | \n",
" state | \n",
" 2014_murders | \n",
" 2015_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" Baltimore | \n",
" Maryland | \n",
" 211 | \n",
" 344 | \n",
" 133 | \n",
"
\n",
" \n",
" 1 | \n",
" Chicago | \n",
" Illinois | \n",
" 411 | \n",
" 478 | \n",
" 67 | \n",
"
\n",
" \n",
" 2 | \n",
" Houston | \n",
" Texas | \n",
" 242 | \n",
" 303 | \n",
" 61 | \n",
"
\n",
" \n",
" 3 | \n",
" Cleveland | \n",
" Ohio | \n",
" 63 | \n",
" 120 | \n",
" 57 | \n",
"
\n",
" \n",
" 4 | \n",
" Washington | \n",
" D.C. | \n",
" 105 | \n",
" 162 | \n",
" 57 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" city state 2014_murders 2015_murders change\n",
"0 Baltimore Maryland 211 344 133\n",
"1 Chicago Illinois 411 478 67\n",
"2 Houston Texas 242 303 61\n",
"3 Cleveland Ohio 63 120 57\n",
"4 Washington D.C. 105 162 57"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2015.head()"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.359791Z",
"start_time": "2018-10-02T04:48:15.261397Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" city | \n",
" state | \n",
" 2015_murders | \n",
" 2016_murders | \n",
" change | \n",
" source | \n",
" as_of | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" Chicago | \n",
" Illinois | \n",
" 378 | \n",
" 536 | \n",
" 158 | \n",
" https://portal.chicagopolice.org/portal/page/p... | \n",
" 10/2/2016 | \n",
"
\n",
" \n",
" 1 | \n",
" Orlando | \n",
" Florida | \n",
" 19 | \n",
" 73 | \n",
" 54 | \n",
" OPD | \n",
" 9/22/2016 | \n",
"
\n",
" \n",
" 2 | \n",
" Memphis | \n",
" Tennessee | \n",
" 114 | \n",
" 158 | \n",
" 44 | \n",
" MPD | \n",
" 9/11/2016 | \n",
"
\n",
" \n",
" 3 | \n",
" Phoenix | \n",
" Arizona | \n",
" 72 | \n",
" 111 | \n",
" 39 | \n",
" PPD | \n",
" 8/31/2016 | \n",
"
\n",
" \n",
" 4 | \n",
" Las Vegas | \n",
" Nevada | \n",
" 90 | \n",
" 125 | \n",
" 35 | \n",
" http://www.lvmpd.com/Sections/Homicide/Homicid... | \n",
" 9/28/2016 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" city state 2015_murders 2016_murders change \\\n",
"0 Chicago Illinois 378 536 158 \n",
"1 Orlando Florida 19 73 54 \n",
"2 Memphis Tennessee 114 158 44 \n",
"3 Phoenix Arizona 72 111 39 \n",
"4 Las Vegas Nevada 90 125 35 \n",
"\n",
" source as_of \n",
"0 https://portal.chicagopolice.org/portal/page/p... 10/2/2016 \n",
"1 OPD 9/22/2016 \n",
"2 MPD 9/11/2016 \n",
"3 PPD 8/31/2016 \n",
"4 http://www.lvmpd.com/Sections/Homicide/Homicid... 9/28/2016 "
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2016.head()"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.452064Z",
"start_time": "2018-10-02T04:48:15.363895Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"2014_murders 6242.776668\n",
"2015_murders 8406.008816\n",
"change 477.806935\n",
"dtype: float64"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2015.var()"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.561832Z",
"start_time": "2018-10-02T04:48:15.457976Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" city | \n",
" state | \n",
" 2015_murders | \n",
" 2016_murders | \n",
" change | \n",
" source | \n",
" as_of | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" Chicago | \n",
" Illinois | \n",
" 378 | \n",
" 536 | \n",
" 158 | \n",
" https://portal.chicagopolice.org/portal/page/p... | \n",
" 10/2/2016 | \n",
"
\n",
" \n",
" 1 | \n",
" Orlando | \n",
" Florida | \n",
" 19 | \n",
" 73 | \n",
" 54 | \n",
" OPD | \n",
" 9/22/2016 | \n",
"
\n",
" \n",
" 2 | \n",
" Memphis | \n",
" Tennessee | \n",
" 114 | \n",
" 158 | \n",
" 44 | \n",
" MPD | \n",
" 9/11/2016 | \n",
"
\n",
" \n",
" 3 | \n",
" Phoenix | \n",
" Arizona | \n",
" 72 | \n",
" 111 | \n",
" 39 | \n",
" PPD | \n",
" 8/31/2016 | \n",
"
\n",
" \n",
" 4 | \n",
" Las Vegas | \n",
" Nevada | \n",
" 90 | \n",
" 125 | \n",
" 35 | \n",
" http://www.lvmpd.com/Sections/Homicide/Homicid... | \n",
" 9/28/2016 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" city state 2015_murders 2016_murders change \\\n",
"0 Chicago Illinois 378 536 158 \n",
"1 Orlando Florida 19 73 54 \n",
"2 Memphis Tennessee 114 158 44 \n",
"3 Phoenix Arizona 72 111 39 \n",
"4 Las Vegas Nevada 90 125 35 \n",
"\n",
" source as_of \n",
"0 https://portal.chicagopolice.org/portal/page/p... 10/2/2016 \n",
"1 OPD 9/22/2016 \n",
"2 MPD 9/11/2016 \n",
"3 PPD 8/31/2016 \n",
"4 http://www.lvmpd.com/Sections/Homicide/Homicid... 9/28/2016 "
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2016.head()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.646928Z",
"start_time": "2018-10-02T04:48:15.568060Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" city | \n",
" state | \n",
" 2014_murders | \n",
" 2015_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" 0 | \n",
" Baltimore | \n",
" Maryland | \n",
" 211 | \n",
" 344 | \n",
" 133 | \n",
"
\n",
" \n",
" 1 | \n",
" Chicago | \n",
" Illinois | \n",
" 411 | \n",
" 478 | \n",
" 67 | \n",
"
\n",
" \n",
" 2 | \n",
" Houston | \n",
" Texas | \n",
" 242 | \n",
" 303 | \n",
" 61 | \n",
"
\n",
" \n",
" 3 | \n",
" Cleveland | \n",
" Ohio | \n",
" 63 | \n",
" 120 | \n",
" 57 | \n",
"
\n",
" \n",
" 4 | \n",
" Washington | \n",
" D.C. | \n",
" 105 | \n",
" 162 | \n",
" 57 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" city state 2014_murders 2015_murders change\n",
"0 Baltimore Maryland 211 344 133\n",
"1 Chicago Illinois 411 478 67\n",
"2 Houston Texas 242 303 61\n",
"3 Cleveland Ohio 63 120 57\n",
"4 Washington D.C. 105 162 57"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2015.head()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.761052Z",
"start_time": "2018-10-02T04:48:15.649036Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/plain": [
"2015_murders 2.338236\n",
"2016_murders 3.194842\n",
"change 4.564821\n",
"dtype: float64"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2016.skew()"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.863101Z",
"start_time": "2018-10-02T04:48:15.767191Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" 2014_murders | \n",
" 2015_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" 2014_murders | \n",
" 1.000000 | \n",
" 0.978106 | \n",
" 0.487938 | \n",
"
\n",
" \n",
" 2015_murders | \n",
" 0.978106 | \n",
" 1.000000 | \n",
" 0.658906 | \n",
"
\n",
" \n",
" change | \n",
" 0.487938 | \n",
" 0.658906 | \n",
" 1.000000 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" 2014_murders 2015_murders change\n",
"2014_murders 1.000000 0.978106 0.487938\n",
"2015_murders 0.978106 1.000000 0.658906\n",
"change 0.487938 0.658906 1.000000"
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2015.corr()"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:15.962201Z",
"start_time": "2018-10-02T04:48:15.870671Z"
},
"scrolled": false
},
"outputs": [
{
"data": {
"text/html": [
"\n",
"\n",
"
\n",
" \n",
" \n",
" | \n",
" 2015_murders | \n",
" 2016_murders | \n",
" change | \n",
"
\n",
" \n",
" \n",
" \n",
" 2015_murders | \n",
" 1.000000 | \n",
" 0.968022 | \n",
" 0.377959 | \n",
"
\n",
" \n",
" 2016_murders | \n",
" 0.968022 | \n",
" 1.000000 | \n",
" 0.598131 | \n",
"
\n",
" \n",
" change | \n",
" 0.377959 | \n",
" 0.598131 | \n",
" 1.000000 | \n",
"
\n",
" \n",
"
\n",
"
"
],
"text/plain": [
" 2015_murders 2016_murders change\n",
"2015_murders 1.000000 0.968022 0.377959\n",
"2016_murders 0.968022 1.000000 0.598131\n",
"change 0.377959 0.598131 1.000000"
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"murder_2016.corr()"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {
"ExecuteTime": {
"end_time": "2018-10-02T04:48:18.379165Z",
"start_time": "2018-10-02T04:48:15.969058Z"
},
"scrolled": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/anand/.virtualenvs/dsutils/lib/python3.6/site-packages/seaborn/axisgrid.py:2262: UserWarning: The `size` paramter has been renamed to `height`; please update your code.\n",
" warnings.warn(msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"# Correlation btw Numerical Columns\n"
]
},
{
"data": {
"text/html": [
"\n",
"\n",
"\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/anand/.virtualenvs/dsutils/lib/python3.6/site-packages/seaborn/axisgrid.py:2262: UserWarning: The `size` paramter has been renamed to `height`; please update your code.\n",
" warnings.warn(msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"# Correlation btw Numerical Columns\n"
]
},
{
"data": {
"text/html": [
"\n",
"\n",
"\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"/home/anand/.virtualenvs/dsutils/lib/python3.6/site-packages/seaborn/axisgrid.py:2262: UserWarning: The `size` paramter has been renamed to `height`; please update your code.\n",
" warnings.warn(msg, UserWarning)\n"
]
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"# Correlation btw Numerical Columns\n"
]
},
{
"data": {
"text/html": [
"\n",
"\n",
"\n"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"image/png": "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\n",
"text/plain": [
"