{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# EDS Case Study\n",
    "\n",
    "Exploring changes in political beliefs\n",
    "\n",
    "Allen Downey\n",
    "\n",
    "[MIT License](https://en.wikipedia.org/wiki/MIT_License)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "from utils import decorate"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loading and validation\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(64814, 105)"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gss = pd.read_hdf('eds.gss.hdf5', 'gss')\n",
    "gss.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def replace_invalid(df):\n",
    "\n",
    "    # recode so depends is in the middle    \n",
    "    d = {1:1, 2:0, 3:0.5}\n",
    "    df.trust.replace(d, inplace=True)\n",
    "    df.helpful.replace(d, inplace=True)\n",
    "    df.racpres.replace(d, inplace=True)\n",
    "    df.owngun.replace(d, inplace=True)\n",
    "    df.fepol.replace(d, inplace=True)\n",
    "    df.sexeduc.replace(d, inplace=True)\n",
    "\n",
    "    d = {1:0, 2:1, 3:0.5}\n",
    "    df.fair.replace(d, inplace=True)\n",
    "\n",
    "replace_invalid(gss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "vars = ['year', 'id_', 'agewed', 'divorce', 'sibs', 'childs', 'age',\n",
    "        'educ', 'paeduc', 'maeduc', 'speduc', 'degree', 'padeg', 'madeg',\n",
    "        'spdeg', 'sex', 'race', 'res16', 'reg16', 'srcbelt', 'partyid',\n",
    "        'pres04', 'pres08', 'pres12', 'polviews', 'natspac', 'natenvir',\n",
    "        'natheal', 'natcity', 'natcrime', 'natdrug', 'nateduc', 'natrace',\n",
    "        'natarms', 'nataid', 'natfare', 'spkath', 'colath', 'libath',\n",
    "        'spkhomo', 'colhomo', 'libhomo', 'cappun', 'gunlaw', 'grass',\n",
    "        'relig', 'fund', 'attend', 'reliten', 'postlife', 'pray',\n",
    "        'relig16', 'fund16', 'sprel16', 'prayer', 'bible', 'racmar',\n",
    "        'racpres', 'affrmact', 'happy', 'hapmar', 'health', 'life',\n",
    "        'helpful', 'fair', 'trust', 'conclerg', 'coneduc', 'confed',\n",
    "        'conpress', 'conjudge', 'conlegis', 'conarmy', 'satjob', 'class_',\n",
    "        'satfin', 'finrela', 'union_', 'fepol', 'abany', 'chldidel',\n",
    "        'sexeduc', 'premarsx', 'xmarsex', 'homosex', 'spanking', 'fear',\n",
    "        'owngun', 'pistol', 'hunt', 'phone', 'memchurh', 'realinc',\n",
    "        'cohort', 'marcohrt', 'ballot', 'wtssall', 'adults', 'compuse',\n",
    "        'databank', 'wtssnr', 'spkrac', 'spkcom', 'spkmil', 'spkmslm']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Validating `polviews`\n",
    "\n",
    "`polviews` contains responses to the [following question](https://gssdataexplorer.norc.org/projects/52787/variables/178/vshow)\n",
    "\n",
    "> We hear a lot of talk these days about liberals and conservatives. \n",
    "I'm going to show you a seven-point scale on which the political views that people might hold are arranged from extremely liberal--point 1--to extremely conservative--point 7. Where would you place yourself on this scale?\n",
    "\n",
    "The following function takes a variable column and returns a series of values and their frequencies."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def values(series):\n",
    "    return series.value_counts().sort_index()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0     1637\n",
       "2.0     6398\n",
       "3.0     6865\n",
       "4.0    21521\n",
       "5.0     8953\n",
       "6.0     8149\n",
       "7.0     1833\n",
       "Name: polviews, dtype: int64"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "values(gss['polviews'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1484"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "gss74 = gss['year']==1974\n",
    "gss74.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.0     21\n",
       "2.0    191\n",
       "3.0    201\n",
       "4.0    543\n",
       "5.0    252\n",
       "6.0    169\n",
       "7.0     34\n",
       "Name: polviews, dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "values(gss.loc[gss74, 'polviews'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<pandas.core.groupby.generic.DataFrameGroupBy object at 0x7ff85732b518>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "by_year = gss.groupby('year')\n",
    "by_year"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "series = by_year['polviews'].mean()\n",
    "series.plot(label='polviews')\n",
    "decorate(ylabel='Mean (7 point scale)',\n",
    "         title='Mean of polviews')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "series = by_year['polviews'].std()\n",
    "series.plot(color='C1', label='polviews')\n",
    "decorate(ylabel='Standard deviation (7 point scale)',\n",
    "         title='Standard deviation of polviews')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Local regression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from statsmodels.nonparametric.smoothers_lowess import lowess\n",
    "\n",
    "def make_lowess(series):\n",
    "    endog = series.values\n",
    "    exog = series.index.values\n",
    "\n",
    "    smooth = lowess(endog, exog)\n",
    "    index, data = np.transpose(smooth)\n",
    "\n",
    "    return pd.Series(data, index=index) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 360x72 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "palette = sns.color_palette('muted', 5)\n",
    "sns.palplot(palette)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "colors = dict(Conservative=palette[3], \n",
    "              Moderate=palette[4], \n",
    "              Liberal=palette[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_series_lowess(series, color):\n",
    "    series.plot(lw=0, marker='o', color=color, alpha=0.5)\n",
    "    smooth = make_lowess(series)\n",
    "    smooth.plot(label='_', color=color)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "series = by_year['polviews'].mean()\n",
    "plot_series_lowess(series, 'C0')\n",
    "decorate(ylabel='Mean (7 point scale)',\n",
    "         title='Mean of polviews')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "series = by_year['polviews'].std()\n",
    "plot_series_lowess(series, color='C1')\n",
    "decorate(ylabel='Standard deviation (7 point scale)',\n",
    "         title='Standard deviation of polviews')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### 3-point scale\n",
    "\n",
    "To make it easier to visualize groups, I'm going to lump the 7-point scale into a 3-point scale.\n",
    "\n",
    "With this scale, there are roughly the same number of people in each group."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Conservative    18935\n",
       "Liberal         14900\n",
       "Moderate        21521\n",
       "Name: polviews3, dtype: int64"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# replace 7 point scale with 3 point scale\n",
    "d = {1:'Liberal', 2:'Liberal', 3:'Liberal', \n",
    "     4:'Moderate', \n",
    "     5:'Conservative', 6:'Conservative', 7:'Conservative'}\n",
    "gss['polviews3'] = gss.polviews.replace(d)\n",
    "values(gss['polviews3'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>polviews3</th>\n",
       "      <th>Conservative</th>\n",
       "      <th>Liberal</th>\n",
       "      <th>Moderate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1974</th>\n",
       "      <td>455</td>\n",
       "      <td>413</td>\n",
       "      <td>543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975</th>\n",
       "      <td>412</td>\n",
       "      <td>424</td>\n",
       "      <td>565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1976</th>\n",
       "      <td>436</td>\n",
       "      <td>389</td>\n",
       "      <td>571</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1977</th>\n",
       "      <td>444</td>\n",
       "      <td>426</td>\n",
       "      <td>579</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1978</th>\n",
       "      <td>454</td>\n",
       "      <td>396</td>\n",
       "      <td>584</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "polviews3  Conservative  Liberal  Moderate\n",
       "year                                      \n",
       "1974                455      413       543\n",
       "1975                412      424       565\n",
       "1976                436      389       571\n",
       "1977                444      426       579\n",
       "1978                454      396       584"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def count_by_year(gss, varname):\n",
    "    \"\"\"\n",
    "    \"\"\"\n",
    "    grouped = gss.groupby([varname, 'year'])\n",
    "    count = grouped[varname].count().unstack(level=0)\n",
    "    \n",
    "    # note: the following is not ideal, because it does not\n",
    "    # distinguish 0 from NA\n",
    "    count = count.replace(0, np.nan).dropna()\n",
    "    return count\n",
    "\n",
    "polviews3_count = count_by_year(gss, 'polviews3')\n",
    "polviews3_count.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "year\n",
       "1974    1411\n",
       "1975    1401\n",
       "1976    1396\n",
       "1977    1449\n",
       "1978    1434\n",
       "dtype: int64"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "total = polviews3_count.sum(axis=1)\n",
    "total.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>polviews3</th>\n",
       "      <th>Conservative</th>\n",
       "      <th>Liberal</th>\n",
       "      <th>Moderate</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>year</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1974</th>\n",
       "      <td>0.322466</td>\n",
       "      <td>0.292700</td>\n",
       "      <td>0.384833</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1975</th>\n",
       "      <td>0.294076</td>\n",
       "      <td>0.302641</td>\n",
       "      <td>0.403283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1976</th>\n",
       "      <td>0.312321</td>\n",
       "      <td>0.278653</td>\n",
       "      <td>0.409026</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1977</th>\n",
       "      <td>0.306418</td>\n",
       "      <td>0.293996</td>\n",
       "      <td>0.399586</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1978</th>\n",
       "      <td>0.316597</td>\n",
       "      <td>0.276151</td>\n",
       "      <td>0.407252</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "polviews3  Conservative   Liberal  Moderate\n",
       "year                                       \n",
       "1974           0.322466  0.292700  0.384833\n",
       "1975           0.294076  0.302641  0.403283\n",
       "1976           0.312321  0.278653  0.409026\n",
       "1977           0.306418  0.293996  0.399586\n",
       "1978           0.316597  0.276151  0.407252"
      ]
     },
     "execution_count": 45,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "polviews3_prop = polviews3_count.div(total, axis=0)\n",
    "polviews3_prop.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_columns_lowess(df, columns, colors):\n",
    "    for col in columns:\n",
    "        series = df[col]\n",
    "        plot_series_lowess(series, colors[col])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "columns = ['Moderate', 'Liberal', 'Conservative']\n",
    "plot_columns_lowess(polviews3_prop, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='Fraction of people with each political view',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Trust"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "trust_mean = gss.groupby('year').trust.mean().dropna()\n",
    "trust_mean.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "plot_series_lowess(trust_mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Fraction',\n",
    "         title='Can people be trusted',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "trust_by_polviews3 = grouped.trust.mean().dropna().unstack(level=0)\n",
    "trust_by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Conservative', 'Moderate']\n",
    "plot_columns_lowess(trust_by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='Can people be trusted?',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fair\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/440/vshow\n",
    "\n",
    "Do you think most people would try to take advantage of you if they got a chance, or would they try to be fair?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fair_mean = gss.groupby('year')['fair'].mean().dropna()\n",
    "plot_series_lowess(fair_mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='Would people try to be fair?',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "fair_by_polviews3 = grouped['fair'].mean().dropna().unstack(level=0)\n",
    "fair_by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(fair_by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='Would people try to be fair?',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## helpful\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/439/vshow\n",
    "\n",
    "Would you say that most of the time people try to be helpful, or that they are mostly just looking out for themselves?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "helpful_mean = gss.groupby('year')['helpful'].mean().dropna()\n",
    "plot_series_lowess(helpful_mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='People try to be helpful?',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "helpful_by_polviews3 = grouped['helpful'].mean().dropna().unstack(level=0)\n",
    "helpful_by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(helpful_by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title='People try to be helpful?',\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## homosex\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/634/vshow\n",
    "\n",
    "What about sexual relations between two adults of the same sex--do you think it is always wrong, almost always wrong, wrong only sometimes, or not wrong at all?\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "varname = 'homosex'\n",
    "title='Sexual relations between adults of the same sex'\n",
    "\n",
    "mean = gss.groupby('year')[varname].mean().dropna()\n",
    "plot_series_lowess(mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Approval',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "by_polviews3 = grouped[varname].mean().dropna().unstack(level=0)\n",
    "by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## racpres\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/400/vshow\n",
    "\n",
    "\n",
    "If your party nominated a (Negro/Black/African-American) for President, would you vote for him if he were qualified for the job?\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "varname = 'racpres'\n",
    "title='Would you vote for a black person for President'\n",
    "\n",
    "mean = gss.groupby('year')[varname].mean().dropna()\n",
    "plot_series_lowess(mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Approval',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "by_polviews3 = grouped[varname].mean().dropna().unstack(level=0)\n",
    "by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## owngun\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/679/vshow\n",
    "\n",
    "Do you happen to have in your home (IF HOUSE: or garage) any guns or revolvers?\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "varname = 'owngun'\n",
    "title='Have gun in home'\n",
    "\n",
    "mean = gss.groupby('year')[varname].mean().dropna()\n",
    "plot_series_lowess(mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Approval',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "by_polviews3 = grouped[varname].mean().dropna().unstack(level=0)\n",
    "by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## fepol\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/591/vshow\n",
    "\n",
    "Tell me if you agree or disagree with this statement: Most men are better suited emotionally for politics than are most women.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "varname = 'fepol'\n",
    "title='Men better suited for politics'\n",
    "\n",
    "mean = gss.groupby('year')[varname].mean().dropna()\n",
    "plot_series_lowess(mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Approval',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "by_polviews3 = grouped[varname].mean().dropna().unstack(level=0)\n",
    "by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Proportion',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## sexeduc\n",
    "\n",
    "https://gssdataexplorer.norc.org/projects/52787/variables/626/vshow\n",
    "\n",
    "Would you be for or against sex education in the public schools?\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "varname = 'sexeduc'\n",
    "title='Favor sex education in the public schools?'\n",
    "\n",
    "mean = gss.groupby('year')[varname].mean().dropna()\n",
    "plot_series_lowess(mean, 'C1')\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Favor',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "by_polviews3 = grouped[varname].mean().dropna().unstack(level=0)\n",
    "by_polviews3.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "columns = ['Liberal', 'Moderate', 'Conservative']\n",
    "plot_columns_lowess(by_polviews3, columns, colors)\n",
    "decorate(xlabel='Year',\n",
    "         ylabel='Favor',\n",
    "         title=title,\n",
    "         xlim=[1970, 2020])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}