05_notifications_per_weekday.ipynb

{
 "cells": [
  {
   "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 matplotlib.patches as patches\n",
    "import os\n",
    "import itertools\n",
    "\n",
    "from multiprocessing import Pool"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%run config.py\n",
    "%run shared.py"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "valid = pd.read_pickle(PATH_VALID_DEVICES)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Notifications per weekday"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**RQ:** How many notifications are in the notification drawer per weekday?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def worker_active_notifications_per_weekday(uuid):\n",
    "    df = pd.read_pickle('%s%s.pkl.gz' % (PATH_DEVICES_DIR, uuid), compression='gzip')\n",
    "    result = pd.DataFrame(columns=[h for h in range(0, 7)])\n",
    "    d = {}\n",
    "    for day in result.columns:\n",
    "        d[day] = df[df.LocalTime.dt.weekday == day].Active.apply(filter_active).apply(len).mean()\n",
    "    return result.append(d, ignore_index=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>3948</th>\n",
       "      <td>0.438819</td>\n",
       "      <td>0.533654</td>\n",
       "      <td>0.423423</td>\n",
       "      <td>0.885593</td>\n",
       "      <td>0.665584</td>\n",
       "      <td>1.070539</td>\n",
       "      <td>0.820000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3949</th>\n",
       "      <td>0.910256</td>\n",
       "      <td>0.636364</td>\n",
       "      <td>1.023256</td>\n",
       "      <td>0.614286</td>\n",
       "      <td>1.252874</td>\n",
       "      <td>0.845161</td>\n",
       "      <td>1.097143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3950</th>\n",
       "      <td>16.761658</td>\n",
       "      <td>17.236486</td>\n",
       "      <td>16.019512</td>\n",
       "      <td>14.595813</td>\n",
       "      <td>16.022075</td>\n",
       "      <td>18.025743</td>\n",
       "      <td>19.060391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3951</th>\n",
       "      <td>1.696154</td>\n",
       "      <td>2.437736</td>\n",
       "      <td>1.996154</td>\n",
       "      <td>1.269388</td>\n",
       "      <td>2.079245</td>\n",
       "      <td>2.160920</td>\n",
       "      <td>2.984375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3952</th>\n",
       "      <td>0.953488</td>\n",
       "      <td>0.295699</td>\n",
       "      <td>0.851351</td>\n",
       "      <td>0.600000</td>\n",
       "      <td>0.397790</td>\n",
       "      <td>1.235955</td>\n",
       "      <td>0.625000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "              0          1          2          3          4          5  \\\n",
       "3948   0.438819   0.533654   0.423423   0.885593   0.665584   1.070539   \n",
       "3949   0.910256   0.636364   1.023256   0.614286   1.252874   0.845161   \n",
       "3950  16.761658  17.236486  16.019512  14.595813  16.022075  18.025743   \n",
       "3951   1.696154   2.437736   1.996154   1.269388   2.079245   2.160920   \n",
       "3952   0.953488   0.295699   0.851351   0.600000   0.397790   1.235955   \n",
       "\n",
       "              6  \n",
       "3948   0.820000  \n",
       "3949   1.097143  \n",
       "3950  19.060391  \n",
       "3951   2.984375  \n",
       "3952   0.625000  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "uuids = valid.UUID.tolist()\n",
    "\n",
    "p = Pool(NUM_CORES)\n",
    "lst = p.map(worker_active_notifications_per_weekday, uuids)\n",
    "\n",
    "weekday = pd.concat(lst)\n",
    "weekday = weekday.reset_index(drop=True)\n",
    "weekday.tail()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "      <td>3953.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>3.357783</td>\n",
       "      <td>3.387484</td>\n",
       "      <td>3.390015</td>\n",
       "      <td>3.453283</td>\n",
       "      <td>3.484108</td>\n",
       "      <td>3.406732</td>\n",
       "      <td>3.315730</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>4.716509</td>\n",
       "      <td>4.624591</td>\n",
       "      <td>4.746775</td>\n",
       "      <td>5.036763</td>\n",
       "      <td>5.113892</td>\n",
       "      <td>4.782215</td>\n",
       "      <td>4.590875</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.184783</td>\n",
       "      <td>1.173333</td>\n",
       "      <td>1.199387</td>\n",
       "      <td>1.205240</td>\n",
       "      <td>1.237113</td>\n",
       "      <td>1.200000</td>\n",
       "      <td>1.151862</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>2.110124</td>\n",
       "      <td>2.154135</td>\n",
       "      <td>2.104478</td>\n",
       "      <td>2.161290</td>\n",
       "      <td>2.184080</td>\n",
       "      <td>2.131234</td>\n",
       "      <td>2.079365</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>3.609375</td>\n",
       "      <td>3.724734</td>\n",
       "      <td>3.725067</td>\n",
       "      <td>3.787234</td>\n",
       "      <td>3.841860</td>\n",
       "      <td>3.812057</td>\n",
       "      <td>3.663551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>59.034483</td>\n",
       "      <td>61.308696</td>\n",
       "      <td>67.847328</td>\n",
       "      <td>103.000000</td>\n",
       "      <td>127.283688</td>\n",
       "      <td>75.337209</td>\n",
       "      <td>61.623037</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 0            1            2            3            4  \\\n",
       "count  3953.000000  3953.000000  3953.000000  3953.000000  3953.000000   \n",
       "mean      3.357783     3.387484     3.390015     3.453283     3.484108   \n",
       "std       4.716509     4.624591     4.746775     5.036763     5.113892   \n",
       "min       0.000000     0.000000     0.000000     0.000000     0.000000   \n",
       "25%       1.184783     1.173333     1.199387     1.205240     1.237113   \n",
       "50%       2.110124     2.154135     2.104478     2.161290     2.184080   \n",
       "75%       3.609375     3.724734     3.725067     3.787234     3.841860   \n",
       "max      59.034483    61.308696    67.847328   103.000000   127.283688   \n",
       "\n",
       "                 5            6  \n",
       "count  3953.000000  3953.000000  \n",
       "mean      3.406732     3.315730  \n",
       "std       4.782215     4.590875  \n",
       "min       0.000000     0.000000  \n",
       "25%       1.200000     1.151862  \n",
       "50%       2.131234     2.079365  \n",
       "75%       3.812057     3.663551  \n",
       "max      75.337209    61.623037  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weekday.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    3.357783\n",
       "1    3.387484\n",
       "2    3.390015\n",
       "3    3.453283\n",
       "4    3.484108\n",
       "5    3.406732\n",
       "6    3.315730\n",
       "dtype: float64"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weekday.mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "6    3.315730\n",
       "0    3.357783\n",
       "1    3.387484\n",
       "2    3.390015\n",
       "5    3.406732\n",
       "3    3.453283\n",
       "4    3.484108\n",
       "dtype: float64"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "weekday.mean().sort_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Notifications per weekday:\n",
      "M  = 3.4\n",
      "SD = 4.8\n",
      "Md = 2.13\n"
     ]
    }
   ],
   "source": [
    "print('Notifications per weekday:')\n",
    "print('M  = %s' % np.round(weekday.mean().mean(), 2))\n",
    "print('SD = %s' % np.round(weekday.std().mean(), 2))\n",
    "print('Md = %s' % np.round(weekday.median().mean(), 2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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": [
    "ax = weekday.fillna(0).boxplot(\n",
    "    showfliers=False,\n",
    "    figsize=(6, 4)\n",
    ")\n",
    "\n",
    "# Ticks\n",
    "ax.set_yticks(range(0, 11, 1))\n",
    "ax.set_yticklabels([x if x % 2 == 0 else '' for x in range(0, 11, 1)])\n",
    "ax.set_xticklabels(['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun'])\n",
    "\n",
    "# Grid\n",
    "ax.set_axisbelow(True)\n",
    "ax.xaxis.grid(alpha=0.0, linestyle='dashed')\n",
    "ax.yaxis.grid(alpha=0.2, linestyle='dashed')\n",
    "\n",
    "# Tweak font sizes\n",
    "#ax.set_xlabel('Mean', fontsize=10)\n",
    "ax.set_ylabel('Mean Notifications (Grouped)', fontsize=10)\n",
    "ax.tick_params(labelsize=10)\n",
    "\n",
    "plt.savefig('figures/mean_active_boxplot_weekdays.png', bbox_inches='tight', pad_inches=0, dpi=300)\n",
    "plt.savefig('figures/mean_active_boxplot_weekdays.pdf', bbox_inches='tight', pad_inches=0)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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": [
    "ax = weekday.fillna(0).mean(axis=0).plot(\n",
    "    kind='line',\n",
    "    xlim=(0, 6),\n",
    "    xticks=range(0, 7, 1),\n",
    "    yticks=range(0, 7, 1),\n",
    "    ylim=(0, 5),\n",
    "    figsize=(6, 4),\n",
    "    legend=True,\n",
    "    grid=True)\n",
    "\n",
    "# Ticks\n",
    "ax.set_xticks(range(0, 7, 1))\n",
    "ax.set_xticklabels(['Mon', 'Tue', 'Wed', 'Thu', 'Fri', 'Sat', 'Sun'])\n",
    "ax.set_yticks(range(0, 6, 1))\n",
    "\n",
    "# Grid\n",
    "ax.set_axisbelow(True)\n",
    "ax.xaxis.grid(alpha=0.2, linestyle='dashed')\n",
    "ax.yaxis.grid(alpha=0.2, linestyle='dashed')\n",
    "\n",
    "# Tweak font sizes\n",
    "ax.set_xlabel('Weekday', fontsize=10)\n",
    "ax.set_ylabel('Mean Notifications (Grouped)', fontsize=10)\n",
    "ax.tick_params(labelsize=10)\n",
    "\n",
    "# Quantiles\n",
    "ax.fill_between(range(0, 7),\n",
    "                [weekday.fillna(0)[x].quantile(.25) for x in weekday.fillna(0).columns],\n",
    "                [weekday.fillna(0)[x].quantile(.75) for x in weekday.fillna(0).columns],\n",
    "                color='#539caf',\n",
    "                alpha=0.2,\n",
    "                zorder=100,\n",
    "                hatch='/')\n",
    "\n",
    "# Std\n",
    "# ax.fill_between(range(0, 24),\n",
    "#                 [weekday.fillna(0)[x].mean() - weekday.fillna(0)[x].std() for x in weekday.fillna(0).columns],\n",
    "#                 [weekday.fillna(0)[x].mean() + weekday.fillna(0)[x].std() for x in weekday.fillna(0).columns],\n",
    "#                 color='#539caf',\n",
    "#                 alpha=0.2)\n",
    "\n",
    "# Overall mean\n",
    "#ax.axhline(weekday.mean().mean(), color='r', linestyle='--')\n",
    "\n",
    "# Legend\n",
    "handles, labels = ax.get_legend_handles_labels()\n",
    "ax.legend(handles=handles+[patches.Patch(color='#539caf', hatch='///', alpha=0.2)],\n",
    "          labels=['Mean', '25th/75th Percentile'],\n",
    "          loc='center',\n",
    "          bbox_to_anchor=(0.5, 1.1),\n",
    "          ncol=3, \n",
    "          prop={'size': 10})\n",
    "\n",
    "plt.savefig('figures/mean_notifications_per_weekday.png', bbox_inches='tight', pad_inches=0, dpi=300)\n",
    "plt.savefig('figures/mean_notifications_per_weekday.pdf', bbox_inches='tight', pad_inches=0)\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "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": "3.6.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}