1-8-Figure 8-Fraction of repeat consumptions of demographic groups.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "from collections import defaultdict\n",
    "from copy import deepcopy\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "\n",
    "from src.config import FIG_DIR, DATA_DIR\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "figure_data_file = os.path.join(FIG_DIR, 'fig_8_data.csv')\n",
    "\n",
    "if not os.path.exists(figure_data_file):\n",
    "\n",
    "    df = pd.read_csv(os.path.join(DATA_DIR, 'MFP.csv'))\n",
    "\n",
    "    def enlist(lst):\n",
    "        \"\"\" Enlist a mixture of list and str, return as original order\"\"\"\n",
    "        res = []\n",
    "        for i in lst:\n",
    "            if isinstance(i, list) or isinstance(i, tuple):\n",
    "                res.extend(i)\n",
    "            else:\n",
    "                if i == i:\n",
    "                    res.append(i)\n",
    "        return res\n",
    "\n",
    "\n",
    "    def user_repeated_cnt_time_bounded(window_size):\n",
    "        user_repeated = defaultdict(int)\n",
    "        for user, t in sorted(dic.keys()):\n",
    "            curr_food = dic[(user, t)]\n",
    "            prev_food = enlist([dic.get((user, n), [])\n",
    "                                for n in np.arange(t - (window_size - 1), t, 1)])\n",
    "            user_repeated[user] += len(set(curr_food) & set(prev_food))\n",
    "        return user_repeated\n",
    "\n",
    "    dic = df.groupby(['uid', 'days'])['food_id'].apply(list).to_dict()\n",
    "\n",
    "    # Or for each user, %repeat consumption = 1- unique/total\n",
    "    df = df.groupby('uid')['food_id'].apply(list).reset_index()\n",
    "    time_bound = 7\n",
    "    user_repeated_cnt = user_repeated_cnt_time_bounded(time_bound)\n",
    "    df['repeated_cnt_' + str(time_bound)] = df['uid'].apply(lambda s: user_repeated_cnt[s])\n",
    "    df['n_consumption'] = df['food_id'].apply(len)\n",
    "    df['n_item'] = df['food_id'].apply(set)\n",
    "    df['n_item'] = df['n_item'].apply(len)\n",
    "    df['perc_repeat'] = 1 - df['n_item'] / df['n_consumption']\n",
    "    df['perc_repeat_7'] = df['repeated_cnt_7'] / df['n_consumption']\n",
    "\n",
    "    # load profile data\n",
    "    file_path = os.path.join(DATA_DIR, 'profiles.csv')\n",
    "    pro_df = pd.read_csv(file_path)\n",
    "\n",
    "    data = pd.merge(df[['uid', 'perc_repeat_7']], pro_df, on='uid', how='left').fillna('NA')\n",
    "\n",
    "    data.to_csv(figure_data_file, index=False)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Plots"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def highest_option(factor):\n",
    "    # at group that is n`ot NA or any\n",
    "    grp_data = data.groupby(factor)['perc_repeat_7'].mean()\n",
    "    return [[l for l in grp_data.sort_values(ascending=False).index.tolist() if l != 'NA'][0]]\n",
    "\n",
    "def plot_fig8(data, cols_ref, context='poster', sharey=True, ncol=3):\n",
    "    # set background param\n",
    "    if context == 'paper':\n",
    "        plt.rc('pdf', fonttype=42)\n",
    "        sns.set(style=\"ticks\")\n",
    "        sns.set_context(context, font_scale=1.5)\n",
    "        figsize = (6, 3)\n",
    "        c1, c2 = 'darkgrey', 'dimgrey'\n",
    "    else:\n",
    "        plt.rc('pdf')\n",
    "        sns.set(style=\"whitegrid\")\n",
    "        sns.set_context(context, font_scale=0.8)\n",
    "        figsize = (12, 6)\n",
    "        c1, c2 = sns.color_palette(\"Paired\")[:2]\n",
    "\n",
    "    fig, axs = plt.subplots(1, ncol, figsize=figsize)\n",
    "\n",
    "    for i, d in enumerate(cols_ref):\n",
    "        factor, xlabel, order, xticks = d\n",
    "        curr_df = data[['perc_repeat_7', factor]]\n",
    "        curr_df2 = deepcopy(curr_df)\n",
    "        curr_df2[factor] = 'Any'\n",
    "        curr_df = curr_df[curr_df[factor] != 'NA']\n",
    "        curr_df = pd.concat([curr_df, curr_df2]).reset_index()\n",
    "\n",
    "        sig_options = highest_option(factor)\n",
    "        my_colors = {o: c1 for o in order}\n",
    "        my_colors.update({o: c2 for o in sig_options})\n",
    "\n",
    "        ax = sns.barplot(x=factor, y='perc_repeat_7', data=curr_df, order=order,\n",
    "                         palette=my_colors, ax=axs[i])\n",
    "        if context == 'paper':\n",
    "            ax.grid(ls=\"dotted\")\n",
    "            ax.set(ylim=(.3, .48))\n",
    "        else:\n",
    "            ax.grid(axis='x', ls='')\n",
    "            ax.set(ylim=(.3, .45))\n",
    "\n",
    "        ax.set_xlabel(xlabel)\n",
    "        ax.set(xticklabels=xticks)\n",
    "\n",
    "        if i == 0:\n",
    "            ax.set_ylabel(\"% Repeat consumption\")\n",
    "        else:\n",
    "            ax.set_ylabel(\"\")\n",
    "            if sharey:\n",
    "                ax.set(yticklabels=[])\n",
    "\n",
    "    sns.despine(offset=10, trim=True)\n",
    "\n",
    "    # save figure\n",
    "    fig.tight_layout()\n",
    "    if context == \"paper\":\n",
    "        figure_name = os.path.join(FIG_DIR, 'paper_fig_8.png')\n",
    "    else:\n",
    "        figure_name = os.path.join(FIG_DIR, 'fig_8.png')\n",
    "    plt.savefig(figure_name, dpi=200, transparent=False, bbox_inches='tight', pad_inches=0.0)\n",
    "    plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = pd.read_csv(figure_data_file)\n",
    "cols_ref = [('gender', 'Gender', ['Male', 'Female'], ['Male', 'Female']),\n",
    "            ('age_group', 'Age groups', ['18-44', '45+'], ['18-44', '45+']),\n",
    "            ('regions_us', 'US Regions', ['Northeast', 'Midwest', 'South', 'West'], ['NE', 'MW', 'S', 'W'])]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x432 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_fig8(data, cols_ref, context='poster')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x216 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_fig8(data, cols_ref, context='paper')"
   ]
  },
  {
   "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}