(M) Analysis - Explore Demographics.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Libraries"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import scipy.stats as stat\n",
    "\n",
    "from math import sqrt\n",
    "from mlgear.utils import show, display_columns\n",
    "from surveyweights import normalize_weights"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Load Processed Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/peterhurford/.virtualenvs/dev/lib/python3.8/site-packages/IPython/core/interactiveshell.py:3145: DtypeWarning: Columns (3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,78,79,80,91,92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  has_raised = await self.run_ast_nodes(code_ast.body, cell_name,\n"
     ]
    }
   ],
   "source": [
    "survey = pd.read_csv('responses_processed_national_weighted.csv').fillna('Not presented')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## 2016-2020 Trump Shifts"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## MALE ##\n",
      "Donald Trump, the Republican    48.866674\n",
      "Joe Biden, the Democrat         51.133326\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       52.488332\n",
      "Hillary Clinton    47.511668\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1407 Weighted N: 542\n",
      "-\n",
      "-\n",
      "## FEMALE ##\n",
      "Donald Trump, the Republican    46.887435\n",
      "Joe Biden, the Democrat         53.112565\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       46.441199\n",
      "Hillary Clinton    53.558801\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1490 Weighted N: 552\n"
     ]
    }
   ],
   "source": [
    "print('## MALE ##')\n",
    "options = ['Joe Biden, the Democrat', 'Donald Trump, the Republican']\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['gender'] == 'Male']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "options2 = ['Hillary Clinton', 'Donald Trump']\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['gender'] == 'Male']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## FEMALE ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['gender'] == 'Female']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['gender'] == 'Female']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## GENDER OF BIDEN SUPPORTERS ##\n",
      "Female               48.522332\n",
      "Male                 50.673269\n",
      "Other                 0.456311\n",
      "Prefer not to say     0.348087\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## GENDER OF TRUMP SUPPORTERS ##\n",
      "Female    46.832101\n",
      "Male      52.945669\n",
      "Other      0.222230\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## GENDER OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gender'].value_counts(normalize=True) * survey_.groupby('gender')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## GENDER OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gender'].value_counts(normalize=True) * survey_.groupby('gender')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## WHITE ##\n",
      "Donald Trump, the Republican    57.60881\n",
      "Joe Biden, the Democrat         42.39119\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       58.691399\n",
      "Hillary Clinton    41.308601\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2074 Weighted N: 792\n",
      "-\n",
      "-\n",
      "## BLACK ##\n",
      "Donald Trump, the Republican    29.884095\n",
      "Joe Biden, the Democrat         70.115905\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       15.86821\n",
      "Hillary Clinton    84.13179\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 260 Weighted N: 102\n",
      "-\n",
      "-\n",
      "## HISPANIC ##\n",
      "Donald Trump, the Republican    29.572799\n",
      "Joe Biden, the Democrat         70.427201\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       37.868872\n",
      "Hillary Clinton    62.131128\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 182 Weighted N: 74\n",
      "-\n",
      "-\n",
      "## ASIAN ##\n",
      "Donald Trump, the Republican    18.112753\n",
      "Joe Biden, the Democrat         81.887247\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       24.389825\n",
      "Hillary Clinton    75.610175\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 357 Weighted N: 149\n"
     ]
    }
   ],
   "source": [
    "print('## WHITE ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'White or Caucasian']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'White or Caucasian']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## BLACK ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Black or African American']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Black or African American']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## HISPANIC ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Hispanic or Latino']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Hispanic or Latino']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## ASIAN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Asian or Asian American']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['race'] == 'Asian or Asian American']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## LATINO MEN ##\n",
      "Donald Trump, the Republican    36.204587\n",
      "Joe Biden, the Democrat         63.795413\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       44.706415\n",
      "Hillary Clinton    55.293585\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 96 Weighted N: 41\n",
      "-\n",
      "-\n",
      "## LATINO WOMEN ##\n",
      "Donald Trump, the Republican    22.649384\n",
      "Joe Biden, the Democrat         77.350616\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       29.026905\n",
      "Hillary Clinton    70.973095\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 84 Weighted N: 31\n"
     ]
    }
   ],
   "source": [
    "print('-')\n",
    "print('-')\n",
    "print('## LATINO MEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Hispanic or Latino') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Hispanic or Latino') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## LATINO WOMEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Hispanic or Latino') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Hispanic or Latino') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## BLACK MEN ##\n",
      "Donald Trump, the Republican    30.973298\n",
      "Joe Biden, the Democrat         69.026702\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       21.731223\n",
      "Hillary Clinton    78.268777\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 106 Weighted N: 40\n",
      "-\n",
      "-\n",
      "## BLACK WOMEN ##\n",
      "Donald Trump, the Republican    27.990895\n",
      "Joe Biden, the Democrat         72.009105\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump        9.639717\n",
      "Hillary Clinton    90.360283\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 153 Weighted N: 61\n"
     ]
    }
   ],
   "source": [
    "print('-')\n",
    "print('-')\n",
    "print('## BLACK MEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Black or African American') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Black or African American') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## BLACK WOMEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Black or African American') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey_['race'] == 'Black or African American') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## RACE OF BIDEN SUPPORTERS ##\n",
      "Asian or Asian American       8.050802\n",
      "Black or African American    15.492874\n",
      "Hispanic or Latino           19.140855\n",
      "Other                         2.501932\n",
      "White or Caucasian           54.813537\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## RACE OF TRUMP SUPPORTERS ##\n",
      "Asian or Asian American       1.946929\n",
      "Black or African American     7.219354\n",
      "Hispanic or Latino            8.787313\n",
      "Other                         0.605254\n",
      "White or Caucasian           81.441149\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## RACE OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['race'].value_counts(normalize=True) * survey_.groupby('race')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## RACE OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['race'].value_counts(normalize=True) * survey_.groupby('race')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## COLLEGE ##\n",
      "Donald Trump, the Republican    41.350876\n",
      "Joe Biden, the Democrat         58.649124\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       45.946087\n",
      "Hillary Clinton    54.053913\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2692 Weighted N: 1079\n",
      "-\n",
      "-\n",
      "## NONCOLLEGE ##\n",
      "Donald Trump, the Republican    59.010903\n",
      "Joe Biden, the Democrat         40.989097\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       57.919898\n",
      "Hillary Clinton    42.080102\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 249 Weighted N: 100\n"
     ]
    }
   ],
   "source": [
    "print('## COLLEGE ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['college']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['college']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## NONCOLLEGE ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[~survey_['college']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[~survey_['college']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## EDUCATION OF BIDEN SUPPORTERS ##\n",
      "Completed graduate school     15.105579\n",
      "Graduated from college        25.989114\n",
      "Graduated from high school    23.045082\n",
      "Less than high school          5.486691\n",
      "Some college, no degree       30.373533\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## EDUCATION OF TRUMP SUPPORTERS ##\n",
      "Completed graduate school     11.685511\n",
      "Graduated from college        18.195844\n",
      "Graduated from high school    27.129215\n",
      "Less than high school         17.779998\n",
      "Some college, no degree       25.209432\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## EDUCATION OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['education'].value_counts(normalize=True) * survey_.groupby('education')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## EDUCATION OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['education'].value_counts(normalize=True) * survey_.groupby('education')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## NONCOLLEGE WHITE ##\n",
      "Donald Trump, the Republican    71.30306\n",
      "Joe Biden, the Democrat         28.69694\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       68.512084\n",
      "Hillary Clinton    31.487916\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 185 Weighted N: 71\n",
      "-\n",
      "-\n",
      "## NON-NCW ##\n",
      "Donald Trump, the Republican    39.837701\n",
      "Joe Biden, the Democrat         60.162299\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       43.798582\n",
      "Hillary Clinton    56.201418\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2756 Weighted N: 1089\n"
     ]
    }
   ],
   "source": [
    "print('## NONCOLLEGE WHITE ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## NON-NCW ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[~survey_['noncollege_white']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[~survey_['noncollege_white']]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## NCW-NESS OF BIDEN SUPPORTERS ##\n",
      "False    86.146431\n",
      "True     13.853569\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## NCW-NESS OF TRUMP SUPPORTERS ##\n",
      "False    62.36629\n",
      "True     37.63371\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## NCW-NESS OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['noncollege_white'].value_counts(normalize=True) * survey_.groupby('noncollege_white')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## NCW-NESS OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['noncollege_white'].value_counts(normalize=True) * survey_.groupby('noncollege_white')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## NONCOLLEGE WHITE MEN ##\n",
      "Donald Trump, the Republican    63.130935\n",
      "Joe Biden, the Democrat         36.869065\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       68.551471\n",
      "Hillary Clinton    31.448529\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 95 Weighted N: 38\n",
      "-\n",
      "-\n",
      "## NONCOLLEGE WHITE WOMEN ##\n",
      "Donald Trump, the Republican    81.764194\n",
      "Joe Biden, the Democrat         18.235806\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       68.645477\n",
      "Hillary Clinton    31.354523\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 87 Weighted N: 33\n"
     ]
    }
   ],
   "source": [
    "print('## NONCOLLEGE WHITE MEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white'] & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white'] & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## NONCOLLEGE WHITE WOMEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white'] & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['noncollege_white'] & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## COLLEGE+ WHITE MEN ##\n",
      "Donald Trump, the Republican    53.047121\n",
      "Joe Biden, the Democrat         46.952879\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       57.353412\n",
      "Hillary Clinton    42.646588\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 918 Weighted N: 386\n",
      "-\n",
      "-\n",
      "## COLLEGE+ WHITE WOMEN ##\n",
      "Donald Trump, the Republican    45.892371\n",
      "Joe Biden, the Democrat         54.107629\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       50.780143\n",
      "Hillary Clinton    49.219857\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 938 Weighted N: 384\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-15-55472f9fdde8>:3: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Male')]\n",
      "<ipython-input-15-55472f9fdde8>:8: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Male')]\n",
      "<ipython-input-15-55472f9fdde8>:19: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Female')]\n",
      "<ipython-input-15-55472f9fdde8>:24: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Female')]\n"
     ]
    }
   ],
   "source": [
    "print('## COLLEGE+ WHITE MEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Male')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## COLLEGE+ WHITE WOMEN ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['college'] & (survey['race'] == 'White or Caucasian') & (survey_['gender'] == 'Female')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## AGES 18-44 ##\n",
      "Donald Trump, the Republican    40.318326\n",
      "Joe Biden, the Democrat         59.681674\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       41.081798\n",
      "Hillary Clinton    58.918202\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2262 Weighted N: 991\n",
      "-\n",
      "-\n",
      "## AGES 45+ ##\n",
      "Donald Trump, the Republican    54.120235\n",
      "Joe Biden, the Democrat         45.879765\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       54.085457\n",
      "Hillary Clinton    45.914543\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 679 Weighted N: 321\n"
     ]
    }
   ],
   "source": [
    "print('## AGES 18-44 ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'].isin(['18-24', '25-44'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'].isin(['18-24', '25-44'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## AGES 45+ ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'].isin(['45-64', '65+'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'].isin(['45-64', '65+'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## AGES 18-24 ##\n",
      "Donald Trump, the Republican    30.833635\n",
      "Joe Biden, the Democrat         69.166365\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       38.285542\n",
      "Hillary Clinton    61.714458\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 284 Weighted N: 132\n",
      "-\n",
      "-\n",
      "## AGES 25-44 ##\n",
      "Donald Trump, the Republican    43.657276\n",
      "Joe Biden, the Democrat         56.342724\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       41.276895\n",
      "Hillary Clinton    58.723105\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1978 Weighted N: 868\n",
      "-\n",
      "-\n",
      "## AGES 45-64 ##\n",
      "Donald Trump, the Republican    50.349118\n",
      "Joe Biden, the Democrat         49.650882\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.142967\n",
      "Hillary Clinton    50.857033\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 563 Weighted N: 271\n",
      "-\n",
      "-\n",
      "## AGES 65+ ##\n",
      "Donald Trump, the Republican    61.183367\n",
      "Joe Biden, the Democrat         38.816633\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       62.816508\n",
      "Hillary Clinton    37.183492\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 116 Weighted N: 63\n"
     ]
    }
   ],
   "source": [
    "print('## AGES 18-24 ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'] == '18-24']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'] == '18-24']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## AGES 25-44 ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'] == '25-44']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'] == '25-44']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## AGES 45-64 ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'] == '45-64']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'] == '45-64']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## AGES 65+ ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['age'] == '65+']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['age'] == '65+']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## AGE OF BIDEN SUPPORTERS ##\n",
      "18-24    15.861644\n",
      "25-44    36.703232\n",
      "45-64    33.466036\n",
      "65+      13.969087\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## AGE OF TRUMP SUPPORTERS ##\n",
      "18-24     7.730734\n",
      "25-44    31.093233\n",
      "45-64    37.103250\n",
      "65+      24.072783\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## AGE OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['age'].value_counts(normalize=True) * survey_.groupby('age')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## AGE OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['age'].value_counts(normalize=True) * survey_.groupby('age')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## UNDER $50K ##\n",
      "Donald Trump, the Republican    48.641944\n",
      "Joe Biden, the Democrat         51.358056\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       48.73304\n",
      "Hillary Clinton    51.26696\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1360 Weighted N: 471\n",
      "-\n",
      "-\n",
      "## ABOVE $50K ##\n",
      "Donald Trump, the Republican    47.280839\n",
      "Joe Biden, the Democrat         52.719161\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.798046\n",
      "Hillary Clinton    50.201954\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1581 Weighted N: 642\n"
     ]
    }
   ],
   "source": [
    "print('## UNDER $50K ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['income'].isin(['Under $15,000', 'Between $15,000 and $49,999'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['income'].isin(['Under $15,000', 'Between $15,000 and $49,999'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## ABOVE $50K ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[~survey_['income'].isin(['Under $15,000', 'Between $15,000 and $49,999'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[~survey_['income'].isin(['Under $15,000', 'Between $15,000 and $49,999'])]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## INCOME OF BIDEN SUPPORTERS ##\n",
      "Between $100,000 and $150,000    14.591811\n",
      "Between $15,000 and $49,999      25.977219\n",
      "Between $50,000 and $74,999      19.249742\n",
      "Between $75,000 and $99,999      10.519130\n",
      "Over $150,000                    20.208583\n",
      "Under $15,000                     9.453515\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## INCOME OF TRUMP SUPPORTERS ##\n",
      "Between $100,000 and $150,000    18.051076\n",
      "Between $15,000 and $49,999      30.367048\n",
      "Between $50,000 and $74,999      18.112227\n",
      "Between $75,000 and $99,999      14.965708\n",
      "Over $150,000                    12.182885\n",
      "Under $15,000                     6.321057\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## INCOME OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['income'].value_counts(normalize=True) * survey_.groupby('income')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## INCOME OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['income'].value_counts(normalize=True) * survey_.groupby('income')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## SOUTH ##\n",
      "Donald Trump, the Republican    63.455881\n",
      "Joe Biden, the Democrat         36.544119\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       63.000635\n",
      "Hillary Clinton    36.999365\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 492 Weighted N: 180\n",
      "-\n",
      "-\n",
      "## NORTHEAST ##\n",
      "Donald Trump, the Republican    49.895102\n",
      "Joe Biden, the Democrat         50.104898\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       47.142046\n",
      "Hillary Clinton    52.857954\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 532 Weighted N: 197\n",
      "-\n",
      "-\n",
      "## PACIFIC ##\n",
      "Donald Trump, the Republican    29.9347\n",
      "Joe Biden, the Democrat         70.0653\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       32.188823\n",
      "Hillary Clinton    67.811177\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 729 Weighted N: 285\n",
      "-\n",
      "-\n",
      "## SOUTHEAST ##\n",
      "Donald Trump, the Republican    45.125679\n",
      "Joe Biden, the Democrat         54.874321\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       60.276913\n",
      "Hillary Clinton    39.723087\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 387 Weighted N: 151\n",
      "-\n",
      "-\n",
      "## MIDWEST ##\n",
      "Donald Trump, the Republican    42.920312\n",
      "Joe Biden, the Democrat         57.079688\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       44.747212\n",
      "Hillary Clinton    55.252788\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 582 Weighted N: 211\n",
      "-\n",
      "-\n",
      "## SOUTHWEST ##\n",
      "Donald Trump, the Republican    42.999545\n",
      "Joe Biden, the Democrat         57.000455\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       54.400936\n",
      "Hillary Clinton    45.599064\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 120 Weighted N: 56\n"
     ]
    }
   ],
   "source": [
    "for region in survey['region'].unique():\n",
    "    if region != 'Other' and region != 'Mountains':\n",
    "        print('-')\n",
    "        print('-')\n",
    "        print('## {} ##'.format(region.upper()))\n",
    "        survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "        survey_ = survey_[survey_['region'] == region]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "        survey_ = survey_[survey_['region'] == region]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "        print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## REGION OF BIDEN SUPPORTERS ##\n",
      "Midwest      24.239625\n",
      "Mountains     3.591881\n",
      "Northeast    16.686010\n",
      "Pacific      20.782407\n",
      "South        14.466905\n",
      "Southeast    13.975096\n",
      "Southwest     6.258076\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## REGION OF TRUMP SUPPORTERS ##\n",
      "Midwest      19.927369\n",
      "Mountains     7.007798\n",
      "Northeast    18.166573\n",
      "Pacific       9.707570\n",
      "South        27.464570\n",
      "Southeast    12.564701\n",
      "Southwest     5.161419\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## REGION OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['region'].value_counts(normalize=True) * survey_.groupby('region')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## REGION OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['region'].value_counts(normalize=True) * survey_.groupby('region')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## RURAL ##\n",
      "Donald Trump, the Republican    65.871112\n",
      "Joe Biden, the Democrat         34.128888\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       59.002653\n",
      "Hillary Clinton    40.997347\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 333 Weighted N: 118\n",
      "-\n",
      "-\n",
      "## URBAN ##\n",
      "Donald Trump, the Republican    43.510449\n",
      "Joe Biden, the Democrat         56.489551\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       44.045349\n",
      "Hillary Clinton    55.954651\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1071 Weighted N: 440\n",
      "-\n",
      "-\n",
      "## SUBURBAN ##\n",
      "Donald Trump, the Republican    45.637235\n",
      "Joe Biden, the Democrat         54.362765\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.693347\n",
      "Hillary Clinton    50.306653\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1537 Weighted N: 559\n"
     ]
    }
   ],
   "source": [
    "for region in survey['urban_rural'].unique():\n",
    "    print('-')\n",
    "    print('-')\n",
    "    print('## {} ##'.format(region.upper()))\n",
    "    survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "    survey_ = survey_[survey_['urban_rural'] == region]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "    survey_ = survey_[survey_['urban_rural'] == region]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "    print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## RURAL MEN ##\n",
      "Donald Trump, the Republican    58.471072\n",
      "Joe Biden, the Democrat         41.528928\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       54.415337\n",
      "Hillary Clinton    45.584663\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 139 Weighted N: 58\n",
      "-\n",
      "-\n",
      "## RURAL WOMEN ##\n",
      "Donald Trump, the Republican    70.106393\n",
      "Joe Biden, the Democrat         29.893607\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       60.899268\n",
      "Hillary Clinton    39.100732\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 193 Weighted N: 63\n",
      "-\n",
      "-\n",
      "## URBAN MEN ##\n",
      "Donald Trump, the Republican    50.03638\n",
      "Joe Biden, the Democrat         49.96362\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       53.113072\n",
      "Hillary Clinton    46.886928\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 565 Weighted N: 241\n",
      "-\n",
      "-\n",
      "## URBAN WOMEN ##\n",
      "Donald Trump, the Republican    35.11727\n",
      "Joe Biden, the Democrat         64.88273\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       32.038141\n",
      "Hillary Clinton    67.961859\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 483 Weighted N: 192\n",
      "-\n",
      "-\n",
      "## SUBURBAN MEN ##\n",
      "Donald Trump, the Republican    46.584741\n",
      "Joe Biden, the Democrat         53.415259\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       51.828237\n",
      "Hillary Clinton    48.171763\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 703 Weighted N: 251\n",
      "-\n",
      "-\n",
      "## SUBURBAN WOMEN ##\n",
      "Donald Trump, the Republican    44.581148\n",
      "Joe Biden, the Democrat         55.418852\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       47.258425\n",
      "Hillary Clinton    52.741575\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 814 Weighted N: 304\n"
     ]
    }
   ],
   "source": [
    "for region in survey['urban_rural'].unique():\n",
    "    print('-')\n",
    "    print('-')\n",
    "    print('## {} MEN ##'.format(region.upper()))\n",
    "    survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "    survey_ = survey_[(survey_['urban_rural'] == region) & (survey_['gender'] == 'Male')]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "    survey_ = survey_[(survey_['urban_rural'] == region) & (survey_['gender'] == 'Male')]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "    print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "    print('-')\n",
    "    print('-')\n",
    "    print('## {} WOMEN ##'.format(region.upper()))\n",
    "    survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "    survey_ = survey_[(survey_['urban_rural'] == region) & (survey_['gender'] == 'Female')]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "    survey_ = survey_[(survey_['urban_rural'] == region) & (survey_['gender'] == 'Female')]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "    print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## URBANNESS OF BIDEN SUPPORTERS ##\n",
      "Rural        9.012434\n",
      "Suburban    57.593993\n",
      "Urban       33.393573\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## URBANNESS OF TRUMP SUPPORTERS ##\n",
      "Rural       19.017689\n",
      "Suburban    52.861289\n",
      "Urban       28.121023\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## URBANNESS OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['urban_rural'].value_counts(normalize=True) * survey_.groupby('urban_rural')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## URBANNESS OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['urban_rural'].value_counts(normalize=True) * survey_.groupby('urban_rural')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## BOOK OF FABLES ##\n",
      "Donald Trump, the Republican    16.066733\n",
      "Joe Biden, the Democrat         83.933267\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       23.163665\n",
      "Hillary Clinton    76.836335\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1399 Weighted N: 570\n",
      "-\n",
      "-\n",
      "## WORD OF GOD ##\n",
      "Donald Trump, the Republican    71.47162\n",
      "Joe Biden, the Democrat         28.52838\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       66.046749\n",
      "Hillary Clinton    33.953251\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 382 Weighted N: 149\n",
      "-\n",
      "-\n",
      "## INSPIRED WORD ##\n",
      "Donald Trump, the Republican    46.425273\n",
      "Joe Biden, the Democrat         53.574727\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       47.601851\n",
      "Hillary Clinton    52.398149\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1038 Weighted N: 447\n"
     ]
    }
   ],
   "source": [
    "for opinion in survey['gss_bible'].unique():\n",
    "    if opinion != 'I don’t know':\n",
    "        print('-')\n",
    "        print('-')\n",
    "        print('## {} ##'.format(opinion.upper()))\n",
    "        survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "        survey_ = survey_[survey_['gss_bible'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "        survey_ = survey_[survey_['gss_bible'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "        print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## BIBLE ORIENTATION OF BIDEN SUPPORTERS ##\n",
      "Book of fables    32.575711\n",
      "I don’t know       4.183576\n",
      "Inspired word     47.309372\n",
      "Word of God       15.931340\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## BIBLE ORIENTATION OF TRUMP SUPPORTERS ##\n",
      "Book of fables     6.817578\n",
      "I don’t know       4.724450\n",
      "Inspired word     44.821301\n",
      "Word of God       43.636670\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## BIBLE ORIENTATION OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_bible'].value_counts(normalize=True) * survey_.groupby('gss_bible')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## BIBLE ORIENTATION OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_bible'].value_counts(normalize=True) * survey_.groupby('gss_bible')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## CAN'T BE TOO CAREFUL ##\n",
      "Donald Trump, the Republican    45.144259\n",
      "Joe Biden, the Democrat         54.855741\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       43.94806\n",
      "Hillary Clinton    56.05194\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1700 Weighted N: 641\n",
      "-\n",
      "-\n",
      "## CAN TRUST ##\n",
      "Donald Trump, the Republican    53.80736\n",
      "Joe Biden, the Democrat         46.19264\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       58.653111\n",
      "Hillary Clinton    41.346889\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1137 Weighted N: 431\n"
     ]
    }
   ],
   "source": [
    "for opinion in survey['gss_trust'].unique():\n",
    "    if opinion != 'Don’t know':\n",
    "        print('-')\n",
    "        print('-')\n",
    "        print('## {} ##'.format(opinion.upper()))\n",
    "        survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "        survey_ = survey_[survey_['gss_trust'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "        survey_ = survey_[survey_['gss_trust'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "        print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## SOCIAL TRUST OF BIDEN SUPPORTERS ##\n",
      "Can trust               30.007481\n",
      "Can't be too careful    65.845324\n",
      "Don’t know               4.147195\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## SOCIAL TRUST OF TRUMP SUPPORTERS ##\n",
      "Can trust               38.215647\n",
      "Can't be too careful    59.244515\n",
      "Don’t know               2.539838\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## SOCIAL TRUST OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_trust'].value_counts(normalize=True) * survey_.groupby('gss_trust')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## SOCIAL TRUST OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_trust'].value_counts(normalize=True) * survey_.groupby('gss_trust')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## PRO-SPANKING ##\n",
      "Donald Trump, the Republican    58.978561\n",
      "Joe Biden, the Democrat         41.021439\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       58.728986\n",
      "Hillary Clinton    41.271014\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1013 Weighted N: 437\n",
      "-\n",
      "-\n",
      "## ANTI-SPANKING ##\n",
      "Donald Trump, the Republican    26.344205\n",
      "Joe Biden, the Democrat         73.655795\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       30.199499\n",
      "Hillary Clinton    69.800501\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1510 Weighted N: 539\n"
     ]
    }
   ],
   "source": [
    "for opinion in survey['gss_spanking'].unique():\n",
    "    if opinion != 'Don\\'t know':\n",
    "        print('-')\n",
    "        print('-')\n",
    "        print('## {} ##'.format('PRO-SPANKING' if opinion == 'Agree' else 'ANTI-SPANKING'))\n",
    "        survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "        survey_ = survey_[survey_['gss_spanking'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "        survey_ = survey_[survey_['gss_spanking'] == opinion]\n",
    "        survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "        print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "        print('-')\n",
    "        weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "        print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## SPANKING ORIENTATION OF BIDEN SUPPORTERS ##\n",
      "Agree         44.510416\n",
      "Disagree      39.729076\n",
      "Don't know    15.760508\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## SPANKING ORIENTATION OF TRUMP SUPPORTERS ##\n",
      "Agree         69.966107\n",
      "Disagree      15.535650\n",
      "Don't know    14.498243\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## SPANKING ORIENTATION OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_spanking'].value_counts(normalize=True) * survey_.groupby('gss_spanking')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## SPANKING ORIENTATION OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['gss_spanking'].value_counts(normalize=True) * survey_.groupby('gss_spanking')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## TWITTER USERS ##\n",
      "Donald Trump, the Republican    49.539308\n",
      "Joe Biden, the Democrat         50.460692\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       55.950378\n",
      "Hillary Clinton    44.049622\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1707 Weighted N: 754\n",
      "-\n",
      "-\n",
      "## NON-TWITTER USERS ##\n",
      "Donald Trump, the Republican    47.246699\n",
      "Joe Biden, the Democrat         52.753301\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       47.655978\n",
      "Hillary Clinton    52.344022\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1234 Weighted N: 516\n"
     ]
    }
   ],
   "source": [
    "for opinion in survey['social_twitter'].unique():\n",
    "    print('-')\n",
    "    print('-')\n",
    "    print('## {} ##'.format('TWITTER USERS' if opinion else 'NON-TWITTER USERS'))\n",
    "    survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "    survey_ = survey_[survey_['social_twitter'] == opinion]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "    survey_ = survey_[survey_['social_twitter'] == opinion]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "    print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## TWITTER USE OF BIDEN SUPPORTERS ##\n",
      "False    77.893842\n",
      "True     22.106158\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## BIBLE ORIENTATION OF TRUMP SUPPORTERS ##\n",
      "False    76.272456\n",
      "True     23.727544\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## TWITTER USE OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['social_twitter'].value_counts(normalize=True) * survey_.groupby('social_twitter')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## BIBLE ORIENTATION OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['social_twitter'].value_counts(normalize=True) * survey_.groupby('social_twitter')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "-\n",
      "-\n",
      "## FB USERS ##\n",
      "Donald Trump, the Republican    48.791952\n",
      "Joe Biden, the Democrat         51.208048\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       51.397938\n",
      "Hillary Clinton    48.602062\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2137 Weighted N: 802\n",
      "-\n",
      "-\n",
      "## NON-FB USERS ##\n",
      "Donald Trump, the Republican    45.339475\n",
      "Joe Biden, the Democrat         54.660525\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       43.706705\n",
      "Hillary Clinton    56.293295\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 804 Weighted N: 306\n"
     ]
    }
   ],
   "source": [
    "for opinion in survey['social_fb'].unique():\n",
    "    print('-')\n",
    "    print('-')\n",
    "    print('## {} ##'.format('FB USERS' if opinion else 'NON-FB USERS'))\n",
    "    survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "    survey_ = survey_[survey_['social_fb'] == opinion]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "    survey_ = survey_[survey_['social_fb'] == opinion]\n",
    "    survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "    print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "    print('-')\n",
    "    weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "    print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## FB USE OF BIDEN SUPPORTERS ##\n",
      "False    30.940438\n",
      "True     69.059562\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## FB USE OF TRUMP SUPPORTERS ##\n",
      "False    28.058982\n",
      "True     71.941018\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## FB USE OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['social_fb'].value_counts(normalize=True) * survey_.groupby('social_fb')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## FB USE OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['social_fb'].value_counts(normalize=True) * survey_.groupby('social_fb')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## ALREADY VOTED ##\n",
      "Donald Trump, the Republican    38.626858\n",
      "Joe Biden, the Democrat         61.373142\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.461175\n",
      "Hillary Clinton    50.538825\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2941 Weighted N: 1107\n",
      "-\n",
      "-\n",
      "## NOT ALREADY VOTED ##\n",
      "Donald Trump, the Republican    54.83259\n",
      "Joe Biden, the Democrat         45.16741\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       55.460117\n",
      "Hillary Clinton    44.539883\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1684 Weighted N: 655\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-36-1236665a4526>:9: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n"
     ]
    }
   ],
   "source": [
    "print('## ALREADY VOTED ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['vote_method_plan'] == 'Not presented']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## NOT ALREADY VOTED ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[survey_['vote_method_plan'] != 'Not presented']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[survey_['vote_method_plan'] != 'Not presented']\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## VOTE BY MAIL ##\n",
      "Donald Trump, the Republican    30.075743\n",
      "Joe Biden, the Democrat         69.924257\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.461175\n",
      "Hillary Clinton    50.538825\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2941 Weighted N: 1107\n",
      "-\n",
      "-\n",
      "## VOTE IN PERSON ##\n",
      "Donald Trump, the Republican    60.285398\n",
      "Joe Biden, the Democrat         39.714602\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       61.196583\n",
      "Hillary Clinton    38.803417\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 1379 Weighted N: 532\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-37-3c61f482f4a3>:3: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[((survey['vote_method'] == 'By mail') |\n",
      "<ipython-input-37-3c61f482f4a3>:11: SettingWithCopyWarning: \n",
      "A value is trying to be set on a copy of a slice from a DataFrame.\n",
      "Try using .loc[row_indexer,col_indexer] = value instead\n",
      "\n",
      "See the caveats in the documentation: http://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html#returning-a-view-versus-a-copy\n",
      "  survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n"
     ]
    }
   ],
   "source": [
    "print('## VOTE BY MAIL ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[((survey['vote_method'] == 'By mail') |\n",
    "                  (survey['vote_method_plan'] == 'By mail, and I’ve already requested and received my mail-in ballot') |\n",
    "                  (survey['vote_method_plan'] == 'By mail, and I’ve already requested my mail-in ballot'))]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## VOTE IN PERSON ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[((survey_['vote_method'] == 'In person') |\n",
    "                  (survey_['vote_method_plan'] == 'In person on election day') |\n",
    "                  (survey_['vote_method_plan'] == 'In person before election day'))]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[((survey_['vote_method'] == 'In person') |\n",
    "                  (survey_['vote_method_plan'] == 'In person on election day') |\n",
    "                  (survey_['vote_method_plan'] == 'In person before election day'))]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## HIGH ELECTION FOCUS ##\n",
      "Donald Trump, the Republican    47.100159\n",
      "Joe Biden, the Democrat         52.899841\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.075859\n",
      "Hillary Clinton    50.924141\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2267 Weighted N: 851\n",
      "-\n",
      "-\n",
      "## LOW ELECTION FOCUS ##\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-38-7877ba1eceb2>:3: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n",
      "<ipython-input-38-7877ba1eceb2>:8: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n",
      "<ipython-input-38-7877ba1eceb2>:19: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[(survey['lv_thought'] != 'Quite a lot')]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Donald Trump, the Republican    49.920991\n",
      "Joe Biden, the Democrat         50.079009\n",
      "dtype: float64\n",
      "-\n",
      "Donald Trump       49.075859\n",
      "Hillary Clinton    50.924141\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 2267 Weighted N: 851\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "<ipython-input-38-7877ba1eceb2>:24: UserWarning: Boolean Series key will be reindexed to match DataFrame index.\n",
      "  survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n"
     ]
    }
   ],
   "source": [
    "print('## HIGH ELECTION FOCUS ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## LOW ELECTION FOCUS ##')\n",
    "survey_ = survey.loc[survey['vote_trump_biden'].isin(options)].copy()\n",
    "survey_ = survey_[(survey['lv_thought'] != 'Quite a lot')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote_trump_biden'].value_counts(normalize=True) * survey_.groupby('vote_trump_biden')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "survey_ = survey.loc[survey['vote2016'].isin(options2)].copy()\n",
    "survey_ = survey_[(survey['lv_thought'] == 'Quite a lot')]\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['vote2016'].value_counts(normalize=True) * survey_.groupby('vote2016')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## ELECTION FOCUS OF BIDEN SUPPORTERS ##\n",
      "Only a little     3.216734\n",
      "Quite a lot      77.188182\n",
      "Some             19.595084\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 3169 Weighted N: 1233\n",
      "-\n",
      "-\n",
      "## ELECTION FOCUS OF TRUMP SUPPORTERS ##\n",
      "Only a little     2.024151\n",
      "Quite a lot      75.138339\n",
      "Some             22.837510\n",
      "dtype: float64\n",
      "-\n",
      "Raw N: 956 Weighted N: 379\n"
     ]
    }
   ],
   "source": [
    "print('## ELECTION FOCUS OF BIDEN SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Joe Biden, the Democrat'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['lv_thought'].value_counts(normalize=True) * survey_.groupby('lv_thought')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## ELECTION FOCUS OF TRUMP SUPPORTERS ##')\n",
    "survey_ = survey[survey['vote_trump_biden'] == 'Donald Trump, the Republican'].copy()\n",
    "survey_['lv_weight_alt'] = normalize_weights(survey_['lv_weight_alt'])\n",
    "print(survey_['lv_thought'].value_counts(normalize=True) * survey_.groupby('lv_thought')['lv_weight_alt'].mean() * 100)\n",
    "print('-')\n",
    "weighted_n = int(np.round(survey_['lv_weight_alt'].apply(lambda w: 1 if w > 1 else w).sum()))\n",
    "print('Raw N: {} Weighted N: {}'.format(len(survey_), weighted_n))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Turnout"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## MALE ##\n",
      "Turnout in poll: 71.0%\n",
      "-\n",
      "-\n",
      "## FEMALE ##\n",
      "Turnout in poll: 65.9%\n"
     ]
    }
   ],
   "source": [
    "print('## MALE ##')\n",
    "survey_ = survey[survey['gender'] == 'Male'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## FEMALE ##')\n",
    "survey_ = survey[survey['gender'] == 'Female'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "## WHITE ##\n",
      "Turnout in poll: 70.2%\n",
      "-\n",
      "-\n",
      "## BLACK ##\n",
      "Turnout in poll: 61.4%\n",
      "-\n",
      "-\n",
      "## HISPANIC ##\n",
      "Turnout in poll: 69.6%\n",
      "-\n",
      "-\n",
      "## ASIAN ##\n",
      "Turnout in poll: 63.9%\n"
     ]
    }
   ],
   "source": [
    "print('## WHITE ##')\n",
    "survey_ = survey[survey['race'] == 'White or Caucasian'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## BLACK ##')\n",
    "survey_ = survey[survey['race'] == 'Black or African American'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## HISPANIC ##')\n",
    "survey_ = survey[survey['race'] == 'Hispanic or Latino'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))\n",
    "\n",
    "print('-')\n",
    "print('-')\n",
    "print('## ASIAN ##')\n",
    "survey_ = survey[survey['race'] == 'Asian or Asian American'].copy()\n",
    "survey_['weight'] = normalize_weights(survey_['weight'])\n",
    "turnout = (survey_['lv_index'] * survey_['weight'] * 100).mean()\n",
    "print('Turnout in poll: {}%'.format(np.round(turnout, 1)))"
   ]
  }
 ],
 "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": "3.8.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}