{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.051494,
     "end_time": "2020-05-09T16:28:35.372952",
     "exception": false,
     "start_time": "2020-05-09T16:28:35.321458",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "# Comparison of Profile Models\n",
    "\n",
    "We sample two profile models:\n",
    "\n",
    "1. The isolated profile model (reported)\n",
    "2. The profile subcomponent of the full model\n",
    "\n",
    "This notebook compares those two models to see what impact full-model training has on user inference."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.04941,
     "end_time": "2020-05-09T16:28:35.473619",
     "exception": false,
     "start_time": "2020-05-09T16:28:35.424209",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "papermill": {
     "duration": 0.067655,
     "end_time": "2020-05-09T16:28:35.591802",
     "exception": false,
     "start_time": "2020-05-09T16:28:35.524147",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "from pathlib import Path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "papermill": {
     "duration": 2.08903,
     "end_time": "2020-05-09T16:28:37.730665",
     "exception": false,
     "start_time": "2020-05-09T16:28:35.641635",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from scipy.special import expit, logit\n",
    "from scipy import stats\n",
    "import seaborn as sns\n",
    "import plotnine as p\n",
    "import matplotlib.pyplot as plt\n",
    "from statsmodels.nonparametric.kde import KDEUnivariate\n",
    "import zarr\n",
    "from IPython.display import display, Markdown"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "papermill": {
     "duration": 0.212284,
     "end_time": "2020-05-09T16:28:38.047309",
     "exception": false,
     "start_time": "2020-05-09T16:28:37.835025",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import bookgender.datatools as dt\n",
    "from bookgender.nbutils import *"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "papermill": {
     "duration": 0.070787,
     "end_time": "2020-05-09T16:28:38.166383",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.095596",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "using figure dir figures/ProfileModelCompare\n"
     ]
    }
   ],
   "source": [
    "fig_dir = init_figs('ProfileModelCompare')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.050194,
     "end_time": "2020-05-09T16:28:38.267301",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.217107",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "papermill": {
     "duration": 0.081597,
     "end_time": "2020-05-09T16:28:38.400784",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.319187",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['AZ', 'BX-E', 'BX-I', 'GR-E', 'GR-I']"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "datasets = list(dt.datasets.keys())\n",
    "datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "papermill": {
     "duration": 0.071584,
     "end_time": "2020-05-09T16:28:38.526096",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.454512",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def load(ds, model):\n",
    "    _zf = zarr.ZipStore(f'data/{ds}/inference/{model}/samples.zarr', mode='r')\n",
    "    _c = zarr.LRUStoreCache(_zf, 2**30)\n",
    "    return zarr.group(_c)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "papermill": {
     "duration": 0.214393,
     "end_time": "2020-05-09T16:28:38.793869",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.579476",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "p_samp = {}\n",
    "f_samp = {}\n",
    "for ds in datasets:\n",
    "    p_samp[ds] = load(ds, 'profile')\n",
    "    f_samp[ds] = load(ds, 'full')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.052264,
     "end_time": "2020-05-09T16:28:38.899895",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.847631",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Compare Summary Parameters"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.051724,
     "end_time": "2020-05-09T16:28:39.003773",
     "exception": false,
     "start_time": "2020-05-09T16:28:38.952049",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "The primary parameters of interest for profiles are $\\mu$ and $\\sigma$ - the mean and variance of the (log odds) proportions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "papermill": {
     "duration": 0.165568,
     "end_time": "2020-05-09T16:28:39.222699",
     "exception": false,
     "start_time": "2020-05-09T16:28:39.057131",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-0.514382</td>\n",
       "      <td>-0.399709</td>\n",
       "      <td>-0.443828</td>\n",
       "      <td>-0.273208</td>\n",
       "      <td>-0.254363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.029801</td>\n",
       "      <td>0.020151</td>\n",
       "      <td>0.017970</td>\n",
       "      <td>0.022427</td>\n",
       "      <td>0.021380</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.618320</td>\n",
       "      <td>-0.470822</td>\n",
       "      <td>-0.512055</td>\n",
       "      <td>-0.366509</td>\n",
       "      <td>-0.349278</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.534410</td>\n",
       "      <td>-0.413218</td>\n",
       "      <td>-0.455956</td>\n",
       "      <td>-0.288445</td>\n",
       "      <td>-0.268497</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-0.514355</td>\n",
       "      <td>-0.399837</td>\n",
       "      <td>-0.443576</td>\n",
       "      <td>-0.273294</td>\n",
       "      <td>-0.254363</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>-0.494401</td>\n",
       "      <td>-0.386026</td>\n",
       "      <td>-0.431779</td>\n",
       "      <td>-0.258152</td>\n",
       "      <td>-0.239769</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>-0.407880</td>\n",
       "      <td>-0.317007</td>\n",
       "      <td>-0.374692</td>\n",
       "      <td>-0.187482</td>\n",
       "      <td>-0.152170</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean      -0.514382     -0.399709     -0.443828     -0.273208     -0.254363\n",
       "std        0.029801      0.020151      0.017970      0.022427      0.021380\n",
       "min       -0.618320     -0.470822     -0.512055     -0.366509     -0.349278\n",
       "25%       -0.534410     -0.413218     -0.455956     -0.288445     -0.268497\n",
       "50%       -0.514355     -0.399837     -0.443576     -0.273294     -0.254363\n",
       "75%       -0.494401     -0.386026     -0.431779     -0.258152     -0.239769\n",
       "max       -0.407880     -0.317007     -0.374692     -0.187482     -0.152170"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_mu = pd.DataFrame(dict((ds, p_samp[ds]['mu']) for ds in datasets))\n",
    "p_mu.index.name = 'Sample'\n",
    "p_mu.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "papermill": {
     "duration": 0.152769,
     "end_time": "2020-05-09T16:28:39.429917",
     "exception": false,
     "start_time": "2020-05-09T16:28:39.277148",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-0.422042</td>\n",
       "      <td>-0.386132</td>\n",
       "      <td>-0.434960</td>\n",
       "      <td>-0.267217</td>\n",
       "      <td>-0.244974</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.027193</td>\n",
       "      <td>0.020111</td>\n",
       "      <td>0.016926</td>\n",
       "      <td>0.022540</td>\n",
       "      <td>0.019699</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-0.529776</td>\n",
       "      <td>-0.465225</td>\n",
       "      <td>-0.499783</td>\n",
       "      <td>-0.352164</td>\n",
       "      <td>-0.315294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-0.439871</td>\n",
       "      <td>-0.399545</td>\n",
       "      <td>-0.446454</td>\n",
       "      <td>-0.282246</td>\n",
       "      <td>-0.258416</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-0.421924</td>\n",
       "      <td>-0.386269</td>\n",
       "      <td>-0.435103</td>\n",
       "      <td>-0.267251</td>\n",
       "      <td>-0.244704</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>-0.403800</td>\n",
       "      <td>-0.372363</td>\n",
       "      <td>-0.423487</td>\n",
       "      <td>-0.252198</td>\n",
       "      <td>-0.231755</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>-0.312990</td>\n",
       "      <td>-0.314502</td>\n",
       "      <td>-0.368429</td>\n",
       "      <td>-0.189616</td>\n",
       "      <td>-0.165044</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean      -0.422042     -0.386132     -0.434960     -0.267217     -0.244974\n",
       "std        0.027193      0.020111      0.016926      0.022540      0.019699\n",
       "min       -0.529776     -0.465225     -0.499783     -0.352164     -0.315294\n",
       "25%       -0.439871     -0.399545     -0.446454     -0.282246     -0.258416\n",
       "50%       -0.421924     -0.386269     -0.435103     -0.267251     -0.244704\n",
       "75%       -0.403800     -0.372363     -0.423487     -0.252198     -0.231755\n",
       "max       -0.312990     -0.314502     -0.368429     -0.189616     -0.165044"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f_mu = pd.DataFrame(dict((ds, f_samp[ds]['mu']) for ds in datasets))\n",
    "f_mu.index.name = 'Sample'\n",
    "f_mu.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "papermill": {
     "duration": 0.919155,
     "end_time": "2020-05-09T16:28:40.404044",
     "exception": false,
     "start_time": "2020-05-09T16:28:39.484889",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc8be239790>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "mu = pd.concat({'Separate': p_mu, 'Full': f_mu}, names=['Model']).reset_index()\n",
    "mu = mu.melt(id_vars=['Model', 'Sample'], var_name='Set')\n",
    "sns.boxplot('Set', 'value', hue='Model', data=mu)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "papermill": {
     "duration": 0.161909,
     "end_time": "2020-05-09T16:28:40.624964",
     "exception": false,
     "start_time": "2020-05-09T16:28:40.463055",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.875013</td>\n",
       "      <td>1.185160</td>\n",
       "      <td>1.083392</td>\n",
       "      <td>1.498967</td>\n",
       "      <td>1.444597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.029923</td>\n",
       "      <td>0.018733</td>\n",
       "      <td>0.017045</td>\n",
       "      <td>0.017982</td>\n",
       "      <td>0.016511</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.755840</td>\n",
       "      <td>1.108950</td>\n",
       "      <td>1.008110</td>\n",
       "      <td>1.433390</td>\n",
       "      <td>1.381690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.854365</td>\n",
       "      <td>1.172610</td>\n",
       "      <td>1.071810</td>\n",
       "      <td>1.486530</td>\n",
       "      <td>1.433578</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.874700</td>\n",
       "      <td>1.185165</td>\n",
       "      <td>1.083350</td>\n",
       "      <td>1.498410</td>\n",
       "      <td>1.444095</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.894970</td>\n",
       "      <td>1.197460</td>\n",
       "      <td>1.094915</td>\n",
       "      <td>1.510945</td>\n",
       "      <td>1.455522</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>2.008700</td>\n",
       "      <td>1.257240</td>\n",
       "      <td>1.146390</td>\n",
       "      <td>1.568860</td>\n",
       "      <td>1.513180</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean       1.875013      1.185160      1.083392      1.498967      1.444597\n",
       "std        0.029923      0.018733      0.017045      0.017982      0.016511\n",
       "min        1.755840      1.108950      1.008110      1.433390      1.381690\n",
       "25%        1.854365      1.172610      1.071810      1.486530      1.433578\n",
       "50%        1.874700      1.185165      1.083350      1.498410      1.444095\n",
       "75%        1.894970      1.197460      1.094915      1.510945      1.455522\n",
       "max        2.008700      1.257240      1.146390      1.568860      1.513180"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_s = pd.DataFrame(dict((ds, p_samp[ds]['sigma']) for ds in datasets))\n",
    "p_s.index.name = 'Sample'\n",
    "p_s.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "papermill": {
     "duration": 0.156828,
     "end_time": "2020-05-09T16:28:40.826409",
     "exception": false,
     "start_time": "2020-05-09T16:28:40.669581",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>1.743360</td>\n",
       "      <td>1.206495</td>\n",
       "      <td>1.073051</td>\n",
       "      <td>1.511378</td>\n",
       "      <td>1.371841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.024639</td>\n",
       "      <td>0.018816</td>\n",
       "      <td>0.014803</td>\n",
       "      <td>0.018005</td>\n",
       "      <td>0.014597</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.652220</td>\n",
       "      <td>1.139190</td>\n",
       "      <td>1.017560</td>\n",
       "      <td>1.445010</td>\n",
       "      <td>1.316520</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.726837</td>\n",
       "      <td>1.193845</td>\n",
       "      <td>1.063130</td>\n",
       "      <td>1.499287</td>\n",
       "      <td>1.362078</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>1.743105</td>\n",
       "      <td>1.206190</td>\n",
       "      <td>1.073090</td>\n",
       "      <td>1.511300</td>\n",
       "      <td>1.371835</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.759890</td>\n",
       "      <td>1.219172</td>\n",
       "      <td>1.082910</td>\n",
       "      <td>1.523432</td>\n",
       "      <td>1.381490</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.824560</td>\n",
       "      <td>1.285000</td>\n",
       "      <td>1.125910</td>\n",
       "      <td>1.577510</td>\n",
       "      <td>1.428880</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean       1.743360      1.206495      1.073051      1.511378      1.371841\n",
       "std        0.024639      0.018816      0.014803      0.018005      0.014597\n",
       "min        1.652220      1.139190      1.017560      1.445010      1.316520\n",
       "25%        1.726837      1.193845      1.063130      1.499287      1.362078\n",
       "50%        1.743105      1.206190      1.073090      1.511300      1.371835\n",
       "75%        1.759890      1.219172      1.082910      1.523432      1.381490\n",
       "max        1.824560      1.285000      1.125910      1.577510      1.428880"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f_s = pd.DataFrame(dict((ds, f_samp[ds]['sigma']) for ds in datasets))\n",
    "f_s.index.name = 'Sample'\n",
    "f_s.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "papermill": {
     "duration": 0.816675,
     "end_time": "2020-05-09T16:28:41.703214",
     "exception": false,
     "start_time": "2020-05-09T16:28:40.886539",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fc8bdeebe50>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "sigma = pd.concat({'Separate': p_s, 'Full': f_s}, names=['Model']).reset_index()\n",
    "sigma = sigma.melt(id_vars=['Model', 'Sample'], var_name='Set')\n",
    "sns.boxplot('Set', 'value', hue='Model', data=sigma)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.06288,
     "end_time": "2020-05-09T16:28:41.828668",
     "exception": false,
     "start_time": "2020-05-09T16:28:41.765788",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Comparing Projected Values\n",
    "\n",
    "Let's now compare the $\\theta$ values - what does each model predict for the distribution of user profile tendencies?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "papermill": {
     "duration": 0.162645,
     "end_time": "2020-05-09T16:28:42.054486",
     "exception": false,
     "start_time": "2020-05-09T16:28:41.891841",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.416734</td>\n",
       "      <td>0.422535</td>\n",
       "      <td>0.412365</td>\n",
       "      <td>0.452728</td>\n",
       "      <td>0.451432</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.299767</td>\n",
       "      <td>0.229406</td>\n",
       "      <td>0.216790</td>\n",
       "      <td>0.269713</td>\n",
       "      <td>0.262539</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000527</td>\n",
       "      <td>0.007993</td>\n",
       "      <td>0.009984</td>\n",
       "      <td>0.002151</td>\n",
       "      <td>0.004607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.139004</td>\n",
       "      <td>0.232165</td>\n",
       "      <td>0.237363</td>\n",
       "      <td>0.216333</td>\n",
       "      <td>0.224470</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.370690</td>\n",
       "      <td>0.402033</td>\n",
       "      <td>0.387333</td>\n",
       "      <td>0.436360</td>\n",
       "      <td>0.433280</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.675581</td>\n",
       "      <td>0.597704</td>\n",
       "      <td>0.572539</td>\n",
       "      <td>0.677640</td>\n",
       "      <td>0.666217</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.998827</td>\n",
       "      <td>0.981249</td>\n",
       "      <td>0.975965</td>\n",
       "      <td>0.997260</td>\n",
       "      <td>0.994236</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean       0.416734      0.422535      0.412365      0.452728      0.451432\n",
       "std        0.299767      0.229406      0.216790      0.269713      0.262539\n",
       "min        0.000527      0.007993      0.009984      0.002151      0.004607\n",
       "25%        0.139004      0.232165      0.237363      0.216333      0.224470\n",
       "50%        0.370690      0.402033      0.387333      0.436360      0.433280\n",
       "75%        0.675581      0.597704      0.572539      0.677640      0.666217\n",
       "max        0.998827      0.981249      0.975965      0.997260      0.994236"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_th = pd.DataFrame(dict((ds, p_samp[ds]['thetaP']) for ds in datasets))\n",
    "p_th.index.name = 'Sample'\n",
    "p_th.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "papermill": {
     "duration": 0.1632,
     "end_time": "2020-05-09T16:28:42.282008",
     "exception": false,
     "start_time": "2020-05-09T16:28:42.118808",
     "status": "completed"
    },
    "tags": []
   },
   "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>AZ</th>\n",
       "      <th>BX-E</th>\n",
       "      <th>BX-I</th>\n",
       "      <th>GR-E</th>\n",
       "      <th>GR-I</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.429019</td>\n",
       "      <td>0.427444</td>\n",
       "      <td>0.411889</td>\n",
       "      <td>0.451355</td>\n",
       "      <td>0.458319</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.289965</td>\n",
       "      <td>0.233149</td>\n",
       "      <td>0.212814</td>\n",
       "      <td>0.270317</td>\n",
       "      <td>0.257499</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.001118</td>\n",
       "      <td>0.008927</td>\n",
       "      <td>0.012457</td>\n",
       "      <td>0.002978</td>\n",
       "      <td>0.003469</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.170024</td>\n",
       "      <td>0.233544</td>\n",
       "      <td>0.238614</td>\n",
       "      <td>0.214576</td>\n",
       "      <td>0.240607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.392241</td>\n",
       "      <td>0.408919</td>\n",
       "      <td>0.392352</td>\n",
       "      <td>0.429607</td>\n",
       "      <td>0.439171</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.676072</td>\n",
       "      <td>0.607898</td>\n",
       "      <td>0.569449</td>\n",
       "      <td>0.681477</td>\n",
       "      <td>0.667624</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>0.996645</td>\n",
       "      <td>0.982007</td>\n",
       "      <td>0.982311</td>\n",
       "      <td>0.994626</td>\n",
       "      <td>0.992481</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AZ          BX-E          BX-I          GR-E          GR-I\n",
       "count  10000.000000  10000.000000  10000.000000  10000.000000  10000.000000\n",
       "mean       0.429019      0.427444      0.411889      0.451355      0.458319\n",
       "std        0.289965      0.233149      0.212814      0.270317      0.257499\n",
       "min        0.001118      0.008927      0.012457      0.002978      0.003469\n",
       "25%        0.170024      0.233544      0.238614      0.214576      0.240607\n",
       "50%        0.392241      0.408919      0.392352      0.429607      0.439171\n",
       "75%        0.676072      0.607898      0.569449      0.681477      0.667624\n",
       "max        0.996645      0.982007      0.982311      0.994626      0.992481"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f_th = pd.DataFrame(dict((ds, f_samp[ds]['thetaP']) for ds in datasets))\n",
    "f_th.index.name = 'Sample'\n",
    "f_th.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "papermill": {
     "duration": 0.123041,
     "end_time": "2020-05-09T16:28:42.470547",
     "exception": false,
     "start_time": "2020-05-09T16:28:42.347506",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "thetaP = pd.concat({'Separate': p_th, 'Full': f_th}, names=['Model']).reset_index()\n",
    "thetaP = thetaP.melt(id_vars=['Model', 'Sample'], var_name='Set')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "papermill": {
     "duration": 3.464608,
     "end_time": "2020-05-09T16:28:46.002095",
     "exception": false,
     "start_time": "2020-05-09T16:28:42.537487",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<seaborn.axisgrid.FacetGrid at 0x7fc8bdbbf810>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 513x1080 with 5 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "grid = sns.FacetGrid(row='Set', hue='Model', data=thetaP, aspect=2)\n",
    "grid.map(sns.kdeplot, 'value')\n",
    "grid.add_legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "papermill": {
     "duration": 7.474995,
     "end_time": "2020-05-09T16:28:53.549256",
     "exception": false,
     "start_time": "2020-05-09T16:28:46.074261",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 640x480 with 5 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<ggplot: (8781259350597)>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(pn.ggplot(thetaP, pn.aes('value', color='Model')) \n",
    " + pn.geom_line(stat='density', adjust=0.5)\n",
    " + pn.facet_grid('Set ~'))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.07705,
     "end_time": "2020-05-09T16:28:53.704862",
     "exception": false,
     "start_time": "2020-05-09T16:28:53.627812",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Individual User Thetas\n",
    "\n",
    "Now let's look at individual user's estimated theta values. How different are they?  We will start by loading each user's posterior expected $\\theta_u$ - by linearity of expectation, the expected difference is the difference in expected values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "papermill": {
     "duration": 10.446951,
     "end_time": "2020-05-09T16:29:04.229896",
     "exception": false,
     "start_time": "2020-05-09T16:28:53.782945",
     "status": "completed"
    },
    "tags": []
   },
   "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></th>\n",
       "      <th>nTheta</th>\n",
       "      <th>Theta</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Set</th>\n",
       "      <th>User</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">AZ</th>\n",
       "      <th>0</th>\n",
       "      <td>1.059797</td>\n",
       "      <td>0.742652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.936946</td>\n",
       "      <td>0.281518</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-2.645083</td>\n",
       "      <td>0.066293</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-1.533403</td>\n",
       "      <td>0.177496</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-2.658765</td>\n",
       "      <td>0.065451</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">GR-I</th>\n",
       "      <th>4995</th>\n",
       "      <td>1.067174</td>\n",
       "      <td>0.744059</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4996</th>\n",
       "      <td>-1.543224</td>\n",
       "      <td>0.176067</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4997</th>\n",
       "      <td>-0.918932</td>\n",
       "      <td>0.285175</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4998</th>\n",
       "      <td>1.510817</td>\n",
       "      <td>0.819182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4999</th>\n",
       "      <td>-2.024523</td>\n",
       "      <td>0.116652</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>25000 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             nTheta     Theta\n",
       "Set  User                    \n",
       "AZ   0     1.059797  0.742652\n",
       "     1    -0.936946  0.281518\n",
       "     2    -2.645083  0.066293\n",
       "     3    -1.533403  0.177496\n",
       "     4    -2.658765  0.065451\n",
       "...             ...       ...\n",
       "GR-I 4995  1.067174  0.744059\n",
       "     4996 -1.543224  0.176067\n",
       "     4997 -0.918932  0.285175\n",
       "     4998  1.510817  0.819182\n",
       "     4999 -2.024523  0.116652\n",
       "\n",
       "[25000 rows x 2 columns]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "p_thu = pd.concat(dict(\n",
    "    (ds, pd.DataFrame({'nTheta': np.mean(p_samp[ds]['nTheta'], axis=0)}))\n",
    "    for ds in datasets\n",
    "), names=['Set', 'User'])\n",
    "p_thu['Theta'] = expit(p_thu['nTheta'])\n",
    "p_thu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "papermill": {
     "duration": 10.527296,
     "end_time": "2020-05-09T16:29:14.842198",
     "exception": false,
     "start_time": "2020-05-09T16:29:04.314902",
     "status": "completed"
    },
    "tags": []
   },
   "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></th>\n",
       "      <th>nTheta</th>\n",
       "      <th>Theta</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>Set</th>\n",
       "      <th>User</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">AZ</th>\n",
       "      <th>0</th>\n",
       "      <td>1.271355</td>\n",
       "      <td>0.780975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.704679</td>\n",
       "      <td>0.330776</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-2.010412</td>\n",
       "      <td>0.118114</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-2.480734</td>\n",
       "      <td>0.077220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-1.701148</td>\n",
       "      <td>0.154315</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th rowspan=\"5\" valign=\"top\">GR-I</th>\n",
       "      <th>4995</th>\n",
       "      <td>1.089918</td>\n",
       "      <td>0.748366</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4996</th>\n",
       "      <td>-1.428438</td>\n",
       "      <td>0.193342</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4997</th>\n",
       "      <td>-0.878642</td>\n",
       "      <td>0.293459</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4998</th>\n",
       "      <td>1.544678</td>\n",
       "      <td>0.824144</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4999</th>\n",
       "      <td>-2.015445</td>\n",
       "      <td>0.117591</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>25000 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "             nTheta     Theta\n",
       "Set  User                    \n",
       "AZ   0     1.271355  0.780975\n",
       "     1    -0.704679  0.330776\n",
       "     2    -2.010412  0.118114\n",
       "     3    -2.480734  0.077220\n",
       "     4    -1.701148  0.154315\n",
       "...             ...       ...\n",
       "GR-I 4995  1.089918  0.748366\n",
       "     4996 -1.428438  0.193342\n",
       "     4997 -0.878642  0.293459\n",
       "     4998  1.544678  0.824144\n",
       "     4999 -2.015445  0.117591\n",
       "\n",
       "[25000 rows x 2 columns]"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "f_thu = pd.concat(dict(\n",
    "    (ds, pd.DataFrame({'nTheta': np.mean(f_samp[ds]['nTheta'], axis=0)}))\n",
    "    for ds in datasets\n",
    "), names=['Set', 'User'])\n",
    "f_thu['Theta'] = expit(f_thu['nTheta'])\n",
    "f_thu"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "papermill": {
     "duration": 0.225292,
     "end_time": "2020-05-09T16:29:15.149844",
     "exception": false,
     "start_time": "2020-05-09T16:29:14.924552",
     "status": "completed"
    },
    "tags": []
   },
   "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>nTheta</th>\n",
       "      <th>Theta</th>\n",
       "      <th>nTheta_f</th>\n",
       "      <th>Theta_f</th>\n",
       "      <th>ndiff</th>\n",
       "      <th>diff</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>25000.000000</td>\n",
       "      <td>25000.000000</td>\n",
       "      <td>25000.000000</td>\n",
       "      <td>25000.000000</td>\n",
       "      <td>25000.000000</td>\n",
       "      <td>25000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>-0.377055</td>\n",
       "      <td>0.425650</td>\n",
       "      <td>-0.351105</td>\n",
       "      <td>0.426963</td>\n",
       "      <td>0.025950</td>\n",
       "      <td>0.001312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>1.310954</td>\n",
       "      <td>0.245157</td>\n",
       "      <td>1.343335</td>\n",
       "      <td>0.247269</td>\n",
       "      <td>0.442744</td>\n",
       "      <td>0.073400</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>-4.611648</td>\n",
       "      <td>0.009838</td>\n",
       "      <td>-4.463211</td>\n",
       "      <td>0.011394</td>\n",
       "      <td>-3.004567</td>\n",
       "      <td>-0.591574</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>-1.302338</td>\n",
       "      <td>0.213772</td>\n",
       "      <td>-1.274608</td>\n",
       "      <td>0.218470</td>\n",
       "      <td>-0.128474</td>\n",
       "      <td>-0.021950</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>-0.409445</td>\n",
       "      <td>0.399045</td>\n",
       "      <td>-0.403842</td>\n",
       "      <td>0.400390</td>\n",
       "      <td>0.002216</td>\n",
       "      <td>0.000395</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.411312</td>\n",
       "      <td>0.601402</td>\n",
       "      <td>0.414463</td>\n",
       "      <td>0.602157</td>\n",
       "      <td>0.165283</td>\n",
       "      <td>0.026456</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>5.259550</td>\n",
       "      <td>0.994829</td>\n",
       "      <td>5.000654</td>\n",
       "      <td>0.993311</td>\n",
       "      <td>3.257669</td>\n",
       "      <td>0.652368</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             nTheta         Theta      nTheta_f       Theta_f         ndiff  \\\n",
       "count  25000.000000  25000.000000  25000.000000  25000.000000  25000.000000   \n",
       "mean      -0.377055      0.425650     -0.351105      0.426963      0.025950   \n",
       "std        1.310954      0.245157      1.343335      0.247269      0.442744   \n",
       "min       -4.611648      0.009838     -4.463211      0.011394     -3.004567   \n",
       "25%       -1.302338      0.213772     -1.274608      0.218470     -0.128474   \n",
       "50%       -0.409445      0.399045     -0.403842      0.400390      0.002216   \n",
       "75%        0.411312      0.601402      0.414463      0.602157      0.165283   \n",
       "max        5.259550      0.994829      5.000654      0.993311      3.257669   \n",
       "\n",
       "               diff  \n",
       "count  25000.000000  \n",
       "mean       0.001312  \n",
       "std        0.073400  \n",
       "min       -0.591574  \n",
       "25%       -0.021950  \n",
       "50%        0.000395  \n",
       "75%        0.026456  \n",
       "max        0.652368  "
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "thetaU = p_thu.join(f_thu, rsuffix='_f')\n",
    "thetaU['ndiff'] = thetaU['nTheta_f'] - thetaU['nTheta']\n",
    "thetaU['diff'] = thetaU['Theta_f'] - thetaU['Theta']\n",
    "thetaU.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "papermill": {
     "duration": 0.110868,
     "end_time": "2020-05-09T16:29:15.343224",
     "exception": false,
     "start_time": "2020-05-09T16:29:15.232356",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.025   -0.167280\n",
       "0.975    0.161608\n",
       "Name: diff, dtype: float64"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "thetaU['diff'].quantile([0.025, 0.975])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "papermill": {
     "duration": 0.116008,
     "end_time": "2020-05-09T16:29:15.542352",
     "exception": false,
     "start_time": "2020-05-09T16:29:15.426344",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    2.500000e+04\n",
       "mean     4.549474e-02\n",
       "std      5.761476e-02\n",
       "min      2.054315e-07\n",
       "25%      7.668859e-03\n",
       "50%      2.407147e-02\n",
       "75%      6.075029e-02\n",
       "max      6.523681e-01\n",
       "Name: diff, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "thetaU['diff'].abs().describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "papermill": {
     "duration": 0.083363,
     "end_time": "2020-05-09T16:29:15.709226",
     "exception": false,
     "start_time": "2020-05-09T16:29:15.625863",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "95%ile?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "papermill": {
     "duration": 0.108652,
     "end_time": "2020-05-09T16:29:15.902028",
     "exception": false,
     "start_time": "2020-05-09T16:29:15.793376",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.16448341395008292"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "thetaU['diff'].abs().quantile(0.95)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "papermill": {
     "duration": 3.278831,
     "end_time": "2020-05-09T16:29:19.265232",
     "exception": false,
     "start_time": "2020-05-09T16:29:15.986401",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x7fc8bd80ea10>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "sns.kdeplot(thetaU['diff'].abs(), cumulative=True)\n",
    "plt.axvline(thetaU['diff'].abs().quantile(0.95), color='grey')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "papermill": {
     "duration": 3.180826,
     "end_time": "2020-05-09T16:29:22.501046",
     "exception": false,
     "start_time": "2020-05-09T16:29:19.320220",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.lines.Line2D at 0x7fc8bc0f37d0>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "sns.kdeplot(thetaU['ndiff'].abs(), cumulative=True)\n",
    "plt.axvline(thetaU['ndiff'].abs().quantile(0.95), color='grey')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "papermill": {
     "duration": 0.075203,
     "end_time": "2020-05-09T16:29:22.668296",
     "exception": false,
     "start_time": "2020-05-09T16:29:22.593093",
     "status": "completed"
    },
    "tags": []
   },
   "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": "3.7.6"
  },
  "papermill": {
   "duration": 49.318824,
   "end_time": "2020-05-09T16:29:23.634157",
   "environment_variables": {},
   "exception": null,
   "input_path": "ProfileModelCompare.ipynb",
   "output_path": "ProfileModelCompare.temp.ipynb",
   "parameters": {},
   "start_time": "2020-05-09T16:28:34.315333",
   "version": "1.2.1"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}