{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "env: MKL_NUM_THREADS=12\n",
      "env: OMP_NUM_THREADS=12\n"
     ]
    }
   ],
   "source": [
    "%env MKL_NUM_THREADS=12\n",
    "%env OMP_NUM_THREADS=12"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import defaultdict\n",
    "\n",
    "import numpy as np\n",
    "import scipy as sp\n",
    "import pandas as pd\n",
    "from ipypb import track\n",
    "\n",
    "from polara.evaluation import evaluation_engine as ee\n",
    "from polara.evaluation.pipelines import random_grid, find_optimal_config\n",
    "from polara.recommender.external.turi.turiwrapper import TuriFactorizationRecommender\n",
    "\n",
    "from data_preprocessing import (get_movielens_data,\n",
    "                                get_bookcrossing_data,\n",
    "                                get_similarity_data,\n",
    "                                prepare_data_model)\n",
    "from utils import (report_results, save_results,\n",
    "                   apply_config, print_data_stats,\n",
    "                   save_training_time, save_cv_training_time)\n",
    "\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "seed = 42"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "experiment_name = 'sgd'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Experiment setup"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_labels = ['ML1M', 'ML10M', 'BX']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# according to https://apple.github.io/turicreate/docs/api/generated/turicreate.recommender.ranking_factorization_recommender.RankingFactorizationRecommender.html\n",
    "init_config = dict(with_data_feedback = False, # implicit case\n",
    "                   ranking_optimization = True,\n",
    "                   solver = 'adagrad',\n",
    "                   sgd_step_size = 0, # let Turi autotune it\n",
    "                   seed = seed,\n",
    "                   max_iterations = 25,\n",
    "                   other_tc_params = {}\n",
    "                   )\n",
    "mf_init_config = dict.fromkeys(data_labels, {'SGD': init_config}) # standard scenario"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "params = {\n",
    "          'regularization': [1e-10, 3e-10, 1e-9, 3e-9, 1e-8, 3e-8, 1e-7, 3e-7, 1e-6, 3e-6],\n",
    "          'linear_regularization': [1e-10, 3e-10, 1e-9, 3e-9, 1e-8, 3e-8, 1e-7, 3e-7, 1e-6, 3e-6],\n",
    "          'rank': [40] # for initial tuning (exploration)\n",
    "         }\n",
    "\n",
    "if init_config['solver'] == 'adagrad':\n",
    "    params.update({\n",
    "                   'adagrad_momentum_weighting': [0.9, 0.95, 0.99]\n",
    "                  })      "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "ranks_grid = [1, 5, 10, 15, 20, 30, 50, 60, 75, 100, 125, 150, 200, 250, 300, 350, 400,\n",
    "              500, 750, 1000, 1250, 1500, 1750, 2000, 2500, 3000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "mf_ranks = {'ML1M': [r for r in ranks_grid if r <= 1000],\n",
    "             'ML10M': [r for r in ranks_grid if r <= 1000],\n",
    "             'BX': [r for r in ranks_grid if r <= 2000]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "topk_values = [1, 3, 10, 20, 30]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "target_metric = 'mrr'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dict = dict.fromkeys(data_labels)\n",
    "meta_dict = dict.fromkeys(data_labels)\n",
    "similarities = dict.fromkeys(data_labels)\n",
    "sim_indices = dict.fromkeys(data_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "all_data = [data_dict, similarities, sim_indices, meta_dict]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Movielens1M"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "lbl = 'ML1M'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dict[lbl], meta_dict[lbl] = get_movielens_data('/data/recsys/movielens/ml-1m.zip',\n",
    "                                                    meta_path='data/meta_info_ml1m.csv',\n",
    "                                                    implicit=True,\n",
    "                                                    filter_no_meta=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# not used actually, simply to onform with general pipeline\n",
    "itemid = meta_dict[lbl].index.name\n",
    "sim_indices[lbl] = {itemid: meta_dict[lbl].index}\n",
    "similarities[lbl] = {itemid: sp.sparse.eye(len(meta_dict[lbl].index))}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Movielens10M"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "lbl = 'ML10M'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_dict[lbl], meta_dict[lbl] = get_movielens_data('/data/recsys/movielens/ml-10m.zip',\n",
    "                                                    meta_path='data/meta_info_ml10m.csv',\n",
    "                                                    implicit=True,\n",
    "                                                    filter_no_meta=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# not used actually, simply to onform with general pipeline\n",
    "itemid = meta_dict[lbl].index.name\n",
    "sim_indices[lbl] = {itemid: meta_dict[lbl].index}\n",
    "similarities[lbl] = {itemid: sp.sparse.eye(len(meta_dict[lbl].index))}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(meta_dict[lbl].applymap(len).sum(axis=1)==0).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## BookCrossing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "lbl = 'BX'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/envs/py36/lib/python3.6/site-packages/pandas/io/parsers.py:1995: DeprecationWarning: invalid escape sequence '\\8'\n",
      "  data = self._reader.read(nrows)\n",
      "/opt/conda/envs/py36/lib/python3.6/site-packages/pandas/io/parsers.py:1995: DeprecationWarning: invalid escape sequence '\\9'\n",
      "  data = self._reader.read(nrows)\n"
     ]
    }
   ],
   "source": [
    "data_dict[lbl], meta_dict[lbl] = get_bookcrossing_data('/data/recsys/bookcrossing/BX-CSV-Dump.zip',\n",
    "                                                       get_books_meta=True,\n",
    "                                                       implicit=True,\n",
    "                                                       pcore=5,\n",
    "                                                       filter_no_meta=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# not used actually, simply to onform with general pipeline\n",
    "itemid = meta_dict[lbl].index.name\n",
    "sim_indices[lbl] = {itemid: meta_dict[lbl].index}\n",
    "similarities[lbl] = {itemid: sp.sparse.eye(len(meta_dict[lbl].index))}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.0"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "(meta_dict[lbl].applymap(len).sum(axis=1)==0).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data stats"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ML1M\n",
      "{'userid': 6038, 'movieid': 3522}\n",
      "density 2.699052132255699\n",
      "similarity matrix density 0.028392958546280524\n",
      "ML10M\n",
      "{'userid': 69797, 'movieid': 10258}\n",
      "density 0.6991397242349022\n",
      "similarity matrix density 0.009748488984207448\n",
      "BX\n",
      "{'userid': 7160, 'isbn': 16273}\n",
      "density 0.18925598044812894\n",
      "similarity matrix density 0.0005841769822585451\n"
     ]
    }
   ],
   "source": [
    "print_data_stats(data_labels, all_data)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Standard experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prepare_recommender_models(data_label, data_models, config):\n",
    "    data_model = data_models[data_label]\n",
    "    mf = TuriFactorizationRecommender(data_model, item_side_info=None)\n",
    "    mf.method = 'SGD'\n",
    "    models = [mf]\n",
    "    apply_config(models, config, data_label)\n",
    "    return models\n",
    "\n",
    "def fine_tune_mf(model, params, label, ntrials=60, record_time_as=None):\n",
    "    param_grid, param_names = random_grid(params, n=ntrials)\n",
    "    best_mf_config, mf_scores = find_optimal_config(model, param_grid, param_names,\n",
    "                                                    target_metric,\n",
    "                                                    return_scores=True,\n",
    "                                                    force_build=True,\n",
    "                                                    iterator=lambda x: track(x, label=label))\n",
    "    model_config = {model.method: dict(zip(param_names, best_mf_config))}\n",
    "    model_scores = {model.method: mf_scores}\n",
    "    try:\n",
    "        if record_time_as:\n",
    "            save_training_time(f'{experiment_name}_{record_time_as}', model, mf_scores.index, label)\n",
    "    finally:\n",
    "        return model_config, model_scores"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## tuning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "config = {}\n",
    "scores = {}\n",
    "times = {}\n",
    "data_models = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'with_data_feedback': False,\n",
       " 'ranking_optimization': True,\n",
       " 'solver': 'adagrad',\n",
       " 'sgd_step_size': 0,\n",
       " 'seed': 42,\n",
       " 'max_iterations': 25,\n",
       " 'other_tc_params': {}}"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "mf_init_config['ML1M']['SGD']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:0; max-width:15ex; vertical-align:middle; text-align:right\"></span>\n",
       "<progress style=\"width:60ex\" max=\"3\" value=\"3\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">3/3</span>\n",
       "<span class=\"Time-label\">[02:08:14<05:13, 2564.75s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[A\u001b[2K\r",
       " [████████████████████████████████████████████████████████████] 3/3 [02:08:14<05:13, 2564.75s/it]\u001b[B"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML1M</span>\n",
       "<progress style=\"width:45ex\" max=\"60\" value=\"60\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">60/60</span>\n",
       "<span class=\"Time-label\">[12:26<00:13, 12.44s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "           ML1M [█████████████████████████████████████████████] 60/60 [12:26<00:13, 12.44s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML10M</span>\n",
       "<progress style=\"width:45ex\" max=\"60\" value=\"60\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">60/60</span>\n",
       "<span class=\"Time-label\">[01:50:25<01:48, 110.42s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "          ML10M [█████████████████████████████████████████████] 60/60 [01:50:25<01:48, 110.42s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">BX</span>\n",
       "<progress style=\"width:45ex\" max=\"60\" value=\"60\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">60/60</span>\n",
       "<span class=\"Time-label\">[05:13<00:05, 5.21s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "             BX [█████████████████████████████████████████████] 60/60 [05:13<00:05, 5.21s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for label in track(data_labels):\n",
    "    data_models[label] = prepare_data_model(label, *all_data, seed)\n",
    "    model, = prepare_recommender_models(label, data_models, mf_init_config)\n",
    "    config[label], scores[label] = fine_tune_mf(model, params, label, ntrials=60, record_time_as='param')\n",
    "del model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/envs/py36/lib/python3.6/site-packages/pandas/plotting/_core.py:1001: UserWarning: Attempted to set non-positive left xlim on a log-scaled axis.\n",
      "Invalid limit will be ignored.\n",
      "  ax.set_xlim(left, right)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "# no meta\n",
    "report_results('tuning', scores);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ML1M': {'SGD': {'regularization': 3e-06,\n",
       "   'linear_regularization': 1e-10,\n",
       "   'rank': 40,\n",
       "   'adagrad_momentum_weighting': 0.95}},\n",
       " 'ML10M': {'SGD': {'regularization': 1e-06,\n",
       "   'linear_regularization': 1e-07,\n",
       "   'rank': 40,\n",
       "   'adagrad_momentum_weighting': 0.99}},\n",
       " 'BX': {'SGD': {'regularization': 3e-06,\n",
       "   'linear_regularization': 1e-08,\n",
       "   'rank': 40,\n",
       "   'adagrad_momentum_weighting': 0.95}}}"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "config"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### saving data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_results(f'{experiment_name}_param', config=config, tuning=scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## rank estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "rank_config = {}\n",
    "rank_scores = {}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:0; max-width:15ex; vertical-align:middle; text-align:right\"></span>\n",
       "<progress style=\"width:60ex\" max=\"3\" value=\"3\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">3/3</span>\n",
       "<span class=\"Time-label\">[02:00:38<10:45, 2412.83s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[A\u001b[2K\r",
       " [████████████████████████████████████████████████████████████] 3/3 [02:00:38<10:45, 2412.83s/it]\u001b[B"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML1M</span>\n",
       "<progress style=\"width:45ex\" max=\"20\" value=\"20\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">20/20</span>\n",
       "<span class=\"Time-label\">[11:07<00:10, 33.33s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "           ML1M [█████████████████████████████████████████████] 20/20 [11:07<00:10, 33.33s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML10M</span>\n",
       "<progress style=\"width:45ex\" max=\"20\" value=\"20\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">20/20</span>\n",
       "<span class=\"Time-label\">[01:38:47<01:33, 296.36s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "          ML10M [█████████████████████████████████████████████] 20/20 [01:38:47<01:33, 296.36s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">BX</span>\n",
       "<progress style=\"width:45ex\" max=\"24\" value=\"24\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">24/24</span>\n",
       "<span class=\"Time-label\">[10:45<00:19, 26.86s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "             BX [█████████████████████████████████████████████] 24/24 [10:45<00:19, 26.86s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for label in track(data_labels):\n",
    "    model, = prepare_recommender_models(label, data_models,\n",
    "                                        [mf_init_config, config]) # initiate with optimal config\n",
    "    rank_config[label], rank_scores[label] = fine_tune_mf(model, {'rank': mf_ranks[label]},\n",
    "                                                          label, ntrials=0, record_time_as='rank')\n",
    "del model"
   ]
  },
  {
   "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"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "# no meta\n",
    "report_results('rank', {lbl: v.sort_index() for lbl, scr in rank_scores.items() for k, v in scr.items()});"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'ML1M': {'SGD': {'rank': 1000}},\n",
       " 'ML10M': {'SGD': {'rank': 750}},\n",
       " 'BX': {'SGD': {'rank': 1500}}}"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rank_config"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### saving data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_results(f'{experiment_name}_rank', config=rank_config, tuning=rank_scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cross-validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:0; max-width:15ex; vertical-align:middle; text-align:right\"></span>\n",
       "<progress style=\"width:60ex\" max=\"3\" value=\"3\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">3/3</span>\n",
       "<span class=\"Time-label\">[01:28:08<06:34, 1762.64s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[A\u001b[2K\r",
       " [████████████████████████████████████████████████████████████] 3/3 [01:28:08<06:34, 1762.64s/it]\u001b[B"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML1M</span>\n",
       "<progress style=\"width:45ex\" max=\"5\" value=\"5\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">5/5</span>\n",
       "<span class=\"Time-label\">[10:20<02:05, 123.91s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "           ML1M [█████████████████████████████████████████████] 5/5 [10:20<02:05, 123.91s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">ML10M</span>\n",
       "<progress style=\"width:45ex\" max=\"5\" value=\"5\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">5/5</span>\n",
       "<span class=\"Time-label\">[01:11:15<13:42, 854.94s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "          ML10M [█████████████████████████████████████████████] 5/5 [01:11:15<13:42, 854.94s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<div><span class=\"Text-label\" style=\"display:inline-block; overflow:hidden; white-space:nowrap; text-overflow:ellipsis; min-width:15ex; max-width:15ex; vertical-align:middle; text-align:right\">BX</span>\n",
       "<progress style=\"width:45ex\" max=\"5\" value=\"5\" class=\"Progress-main\"/></progress>\n",
       "<span class=\"Progress-label\"><strong>100%</strong></span>\n",
       "<span class=\"Iteration-label\">5/5</span>\n",
       "<span class=\"Time-label\">[06:34<01:20, 78.72s/it]</span></div>"
      ],
      "text/plain": [
       "\u001b[A\u001b[2K\r",
       "             BX [█████████████████████████████████████████████] 5/5 [06:34<01:20, 78.72s/it]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result = {}\n",
    "for label in track(data_labels):\n",
    "    models = prepare_recommender_models(label, data_models, [mf_init_config, config, rank_config])\n",
    "    result[label] = ee.run_cv_experiment(models,\n",
    "                                         fold_experiment=ee.topk_test,\n",
    "                                         topk_list=topk_values,\n",
    "                                         ignore_feedback=True,\n",
    "                                         iterator=lambda x: track(x, label=label))\n",
    "    save_cv_training_time(experiment_name, models, label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEWCAYAAACKSkfIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAFshJREFUeJzt3X+QVeWd5/H3lwbBH0QN4NZG0MaRJKI4GJG4G9dQskN03RVnV0swG4llSlMOqaTMVJZsEkXHGNxN4m7VOlWS0YRojBqjO1TEtYyOujFGBSUYgm5agtKjRgRkdCIi8N0/7kE61277NDR9G573q6qrz3nOc8793lN9P/f0uec+JzITSVIZhrS6AEnSwDH0Jakghr4kFcTQl6SCGPqSVBBDX5IKYuhLUkEMfRUjItZExJaIGN3UvjwiMiLaI+IHEXF1D+v/TUQ8ExFbI2J+07LPVtv4blP72VX7D/r7+Ui7wtBXaX4PzN4xExGTgP1rrtsBfAW4p4flzwPnRcTQLm0XAP9vF+qU9ghDX6W5mUYQ7zAH+GGdFTNzUWbeC7zRQ5dXgGeATwFExAeBfw0s3uVqpX5m6Ks0vwI+EBHHREQbcB5wSz9u/4fsfFOZBfw98HY/bl/aLYa+SrTjaP8vgGeBf+zHbd8NTIuIg6vHqPVfhDRQhvbeRdrn3Aw8Aoynn0M5M9+KiHuArwOjM/PRiDijPx9D2h0e6as4mfkCjQ90/x1w1x54iB8CX6bx5iINKh7pq1QXAYdm5j83XW0D0BYRI7rMb8/MLRExDGijcbA0tOrzTmZua1r/YRqnjp7eU8VLu8ojfRUpM5/PzKU9LJ4HvNXl58Gq/XvV/Gzga9X0Z7rZdmbmA5m5od8Ll3ZTeBMVSSqHR/qSVBBDX5IKYuhLUkEMfUkqyKC7ZHP06NHZ3t7e6jIkaa+ybNmy1zJzTG/9aoV+RJwO/E8a1yj/XWYuaFp+KvA/gOOBWZl5Z5dlc2h8OxHg6sxc9H6P1d7eztKlPV1JJ0nqTkS8UKdfr6d3qkGprgfOACYCsyNiYlO3F4HPArc2rftB4Arg48BU4IqIOLROYZKk/lfnnP5UoCMzV2fmFuA2YGbXDpm5JjNXANub1v0UcH9mbsjMjcD9wOn9ULckaRfUCf3DgbVd5jurtjpqrRsRF0fE0ohYum7dupqbliT1VZ1z+tFNW92v8dZaNzMXAgsBpkyZ8p7l77zzDp2dnWzevLnmw+59RowYwdixYxk2bFirS5G0D6sT+p3AuC7zY4GXam6/E5jWtO5DNdfduZHOTkaOHEl7ezsR3b2P7N0yk/Xr19PZ2cn48eNbXY6kfVid0ztPAhMiYnxE7EfjbkB1b/92HzAjIg6tPsCdUbX1yebNmxk1atQ+GfgAEcGoUaP26f9kJA0OvYZ+Zm4F5tII61XAHZm5MiKuioizACLipIjoBM4FboiIldW6G4C/ofHG8SRw1a6OPLivBv4O+/rzkzQ41LpOPzOXAEua2i7vMv0kjVM33a17E3DTbtQoSeonDsPQB+3t7bz22mu73UeSWmXQDcMgSQOpfd49rS6BNQvOHLDH2ueP9NesWcNHP/pRPve5z3Hcccfx6U9/mp///Od84hOfYMKECTzxxBNs2LCBs88+m+OPP56TTz6ZFStWALB+/XpmzJjBCSecwCWXXELXG87ccsstTJ06lcmTJ3PJJZewbVvzHfMkafDZ50MfoKOjgy9+8YusWLGCZ599lltvvZVf/OIXfPvb3+aaa67hiiuu4IQTTmDFihVcc801XHDBBQBceeWVnHLKKTz99NOcddZZvPjiiwCsWrWK22+/nUcffZTly5fT1tbGj370o1Y+RUmqpYjTO+PHj2fSpEkAHHvssUyfPp2IYNKkSaxZs4YXXniBn/70pwCcdtpprF+/nk2bNvHII49w1113AXDmmWdy6KGNYYMeeOABli1bxkknnQTAW2+9xWGHHdaCZyZJfVNE6A8fPvzd6SFDhrw7P2TIELZu3crQoe/dDTsuoezuUsrMZM6cOXzrW9/aQxVL0p5RxOmd3px66qnvnp556KGHGD16NB/4wAf+pP3ee+9l48aNAEyfPp0777yTV199FYANGzbwwgu1RjWVpJYq4ki/N/Pnz+fCCy/k+OOP54ADDmDRosaQ/1dccQWzZ8/mYx/7GJ/85Cc54ogjAJg4cSJXX301M2bMYPv27QwbNozrr7+eI488spVPQ6qttCtWtFN0vSJlMJgyZUo230Rl1apVHHPMMS2qaOCU8jzVeob+TvvKvoiIZZk5pbd+nt6RpIIY+pJUEENfkgpi6EtSQQx9SSqIoS9JBdkrr9Pv70us6l4u9c1vfpNbb72VtrY2hgwZwg033MCJJ57I5Zdfzk9+8hMOPPBAAM4991y+9rWvAdDW1sakSZN45513GDp0KHPmzOFLX/oSQ4b4fitp4O2Vod8Kjz32GD/72c946qmnGD58OK+99hpbtmzh61//Oq+88grPPPMMI0aM4I033uA73/nOu+vtv//+LF++HIBXX32V888/n02bNnHllVe26qlIKpihX9PLL7/M6NGj3x23Z/To0fzxj3/ke9/7HmvWrGHEiBEAjBw5kvnz53e7jcMOO4yFCxdy0kknMX/+fG+RKGnAeY6hphkzZrB27Vo+/OEPc+mll/Lwww/T0dHBEUccwciRI2tv56ijjmL79u3vjtsjSQPJ0K/poIMOYtmyZSxcuJAxY8Zw3nnn8dBDD/1Jn+9///tMnjyZcePGsXbt2h63NdiGvpBUDk/v9EFbWxvTpk1j2rRpTJo0iRtuuIEXX3yRN954g5EjR3LhhRdy4YUXctxxx/V4J63Vq1fT1tbm+PstsK+MsSLtDo/0a3ruuef43e9+9+788uXL+chHPsJFF13E3Llz2bx5MwDbtm1jy5Yt3W5j3bp1fP7zn2fu3Lmez5fUEnvlkX4rjpbefPNNvvCFL/D6668zdOhQjj76aBYuXMjBBx/MN77xDY477jhGjhzJ/vvvz5w5c/jQhz4ENO6qNXny5Hcv2fzMZz7DZZddNuD1SxLspaHfCieeeCK//OUvu122YMECFixY0O0yb5guaTDx9I4kFcTQl6SC7DWhv69f5rivPz9Jg8NeEfojRoxg/fr1+2wwZibr169/91u9krSn7BUf5I4dO5bOzk7WrVvX6lL2mBEjRjB27NhWlyFpH7dXhP6wYcMYP358q8uQpL3eXnF6R5LUPwx9SSqIoS9JBTH0JakgtUI/Ik6PiOcioiMi5nWzfHhE3F4tfzwi2qv2YRGxKCKeiYhVEfHV/i1fktQXvYZ+RLQB1wNnABOB2RExsanbRcDGzDwauA64tmo/FxiemZOAE4FLdrwhSJIGXp0j/alAR2auzswtwG3AzKY+M4FF1fSdwPRojB2cwIERMRTYH9gC/FO/VC5J6rM6oX840PU2UJ1VW7d9MnMrsAkYReMN4J+Bl4EXgW9n5obmB4iIiyNiaUQs3Ze/gCVJrVYn9Lu720fzeAg99ZkKbAM+BIwHvhwRR72nY+bCzJySmVPGjBlToyRJ0q6oE/qdwLgu82OBl3rqU53KORjYAJwP/J/MfCczXwUeBabsbtGSpF1TJ/SfBCZExPiI2A+YBSxu6rMYmFNNnwM8mI3R0V4ETouGA4GTgWf7p3RJUl/1GvrVOfq5wH3AKuCOzFwZEVdFxFlVtxuBURHRAVwG7Lis83rgIOA3NN48vp+ZK/r5OUiSaqo14FpmLgGWNLVd3mV6M43LM5vXe7O7dklSa/iNXEkqiKEvSQUx9CWpIIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVxNCXpIIY+pJUEENfkgpSazx97b3a593T6hIAWLPgzFaXIAmP9CWpKIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVxNCXpIIY+pJUEENfkgpi6EtSQQx9SSqIoS9JBTH0Jakghr4kFaRW6EfE6RHxXER0RMS8bpYPj4jbq+WPR0R7l2XHR8RjEbEyIp6JiBH9V74kqS96Df2IaAOuB84AJgKzI2JiU7eLgI2ZeTRwHXBtte5Q4Bbg85l5LDANeKffqpck9UmdI/2pQEdmrs7MLcBtwMymPjOBRdX0ncD0iAhgBrAiM38NkJnrM3Nb/5QuSeqrOqF/OLC2y3xn1dZtn8zcCmwCRgEfBjIi7ouIpyLiK909QERcHBFLI2LpunXr+vocJEk11Qn96KYta/YZCpwCfLr6/ZcRMf09HTMXZuaUzJwyZsyYGiVJknZFndDvBMZ1mR8LvNRTn+o8/sHAhqr94cx8LTP/CCwBPra7RUuSdk2d0H8SmBAR4yNiP2AWsLipz2JgTjV9DvBgZiZwH3B8RBxQvRl8Evht/5QuSeqrob11yMytETGXRoC3ATdl5sqIuApYmpmLgRuBmyOig8YR/qxq3Y0R8V0abxwJLMnMe/bQc5Ek9aLX0AfIzCU0Ts10bbu8y/Rm4Nwe1r2FxmWbkqQW8xu5klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVxNCXpIIY+pJUEENfkgpi6EtSQQx9SSqIoS9JBTH0Jakghr4kFcTQl6SCGPqSVBBDX5IKYuhLUkEMfUkqSK0bo+9t2ufd0+oSAFiz4MxWlyBJf8IjfUkqiKEvSQUx9CWpIIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVpFboR8TpEfFcRHRExLxulg+PiNur5Y9HRHvT8iMi4s2I+Ov+KVuStCt6Df2IaAOuB84AJgKzI2JiU7eLgI2ZeTRwHXBt0/LrgHt3v1xJ0u6oc6Q/FejIzNWZuQW4DZjZ1GcmsKiavhOYHhEBEBFnA6uBlf1TsiRpV9UJ/cOBtV3mO6u2bvtk5lZgEzAqIg4E/gtw5fs9QERcHBFLI2LpunXr6tYuSeqjOqEf3bRlzT5XAtdl5pvv9wCZuTAzp2TmlDFjxtQoSZK0K+rcOasTGNdlfizwUg99OiNiKHAwsAH4OHBORPw34BBge0Rszsz/tduVS5L6rE7oPwlMiIjxwD8Cs4Dzm/osBuYAjwHnAA9mZgL/ZkeHiJgPvGngS1Lr9Br6mbk1IuYC9wFtwE2ZuTIirgKWZuZi4Ebg5ojooHGEP2tPFi1J2jW1boyemUuAJU1tl3eZ3gyc28s25u9CfZKkfuQ3ciWpIIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVxNCXpIIY+pJUEENfkgpi6EtSQQx9SSqIoS9JBTH0Jakghr4kFcTQl6SCGPqSVBBDX5IKYuhLUkEMfUkqiKEvSQUx9CWpIIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVpFboR8TpEfFcRHRExLxulg+PiNur5Y9HRHvV/hcRsSwinql+n9a/5UuS+qLX0I+INuB64AxgIjA7IiY2dbsI2JiZRwPXAddW7a8B/yEzJwFzgJv7q3BJUt/VOdKfCnRk5urM3ALcBsxs6jMTWFRN3wlMj4jIzKcz86WqfSUwIiKG90fhkqS+qxP6hwNru8x3Vm3d9snMrcAmYFRTn/8EPJ2Zbzc/QERcHBFLI2LpunXr6tYuSeqjOqEf3bRlX/pExLE0Tvlc0t0DZObCzJySmVPGjBlToyRJ0q6oE/qdwLgu82OBl3rqExFDgYOBDdX8WOBu4ILMfH53C5Yk7bo6of8kMCEixkfEfsAsYHFTn8U0PqgFOAd4MDMzIg4B7gG+mpmP9lfRkqRd02voV+fo5wL3AauAOzJzZURcFRFnVd1uBEZFRAdwGbDjss65wNHANyJiefVzWL8/C0lSLUPrdMrMJcCSprbLu0xvBs7tZr2rgat3s0ZJUj/xG7mSVBBDX5IKYuhLUkEMfUkqiKEvSQUx9CWpIIa+JBXE0Jekghj6klQQQ1+SCmLoS1JBDH1JKoihL0kFMfQlqSCGviQVxNCXpIIY+pJUEENfkgpi6EtSQQx9SSqIoS9JBTH0Jakghr4kFcTQl6SCGPqSVBBDX5IKYuhLUkEMfUkqiKEvSQUx9CWpIIa+JBXE0Jekghj6klQQQ1+SClIr9CPi9Ih4LiI6ImJeN8uHR8Tt1fLHI6K9y7KvVu3PRcSn+q90SVJf9Rr6EdEGXA+cAUwEZkfExKZuFwEbM/No4Drg2mrdicAs4FjgdOBvq+1JklqgzpH+VKAjM1dn5hbgNmBmU5+ZwKJq+k5gekRE1X5bZr6dmb8HOqrtSZJaYGiNPocDa7vMdwIf76lPZm6NiE3AqKr9V03rHt78ABFxMXBxNftmRDxXq/o9azTw2u5sIK7tp0paz32x027ti31oP4D7oqvBsC+OrNOpTuhHN21Zs0+ddcnMhcDCGrUMmIhYmplTWl3HYOC+2Ml9sZP7Yqe9aV/UOb3TCYzrMj8WeKmnPhExFDgY2FBzXUnSAKkT+k8CEyJifETsR+OD2cVNfRYDc6rpc4AHMzOr9lnV1T3jgQnAE/1TuiSpr3o9vVOdo58L3Ae0ATdl5sqIuApYmpmLgRuBmyOig8YR/qxq3ZURcQfwW2Ar8FeZuW0PPZf+NqhON7WY+2In98VO7oud9pp9EY0DcklSCfxGriQVxNCXpIIY+k0i4qaIeDUiftPqWlopIkZExBMR8euIWBkRV7a6poHW3d9CRHwwIu6PiN9Vvw9tZY0DISLGRcQ/RMSq6m/hi1V7ifui29dFdaHL49W+uL266GVQMvTf6wc0howo3dvAaZn558Bk4PSIOLnFNQ20H/Dev4V5wAOZOQF4oJrf120FvpyZxwAnA39VDbFS4r7o6XVxLXBdtS820hiaZlAy9Jtk5iM0rkAqWja8Wc0Oq36K+tS/h7+FrkOOLALOHtCiWiAzX87Mp6rpN4BVNL5ZX+K+6Ol1cRqNIWhgkO8LQ189ioi2iFgOvArcn5mPt7qmQeBfZObL0AhD4LAW1zOgqhF0TwAep9B90fy6AJ4HXs/MrVWXboebGSwMffUoM7dl5mQa36SeGhHHtbomtU5EHAT8FPhSZv5Tq+tplebXBXBMd90Gtqr6DH31KjNfBx7CzzoA/hAR/xKg+v1qi+sZEBExjEbg/ygz76qai9wXO3R5XZwMHFINQQODfLgZQ1/diogxEXFINb0/8G+BZ1tb1aDQdciROcDft7CWAVENk34jsCozv9tlUYn7orvXxSrgH2gMQQODfF/4jdwmEfFjYBqNoVL/AFyRmTe2tKgWiIjjaXwg1Ubj4OCOzLyqtVUNrO7+FoD/DdwBHAG8CJybmfv0B/8RcQrwf4FngO1V83+lcV6/tH3R7esiIo6ica+RDwJPA/85M99uXaU9M/QlqSCe3pGkghj6klQQQ1+SCmLoS1JBDH1JKoihr2JFxCERcWmr65AGkqGvkh0CGPoqSq/3yJX2YQuAP6sGz7q/ajuDxrgpV2fm7RExDbgKWA98BHgEuDQzt3fdUER8FjgLOAD4M+DuzPzKQDwJqS880lfJ5gHPV4Nn/YrG+Oh/TuOr9f99x7gyNAbV+jIwiUag/8cetjcZOK/qd15EjNuDtUu7xNCXGk4BflyNoPgH4GHgpGrZE5m5OjO3AT+u+nbngczclJmbgd8CR+7xqqU+MvSlhnifZc1jlWRE/GVELK9+plTtXcda2YanTzUIGfoq2RvAyGr6ERqnZNoiYgxwKvBEtWxqdQ/UITRO3/wiM+/OzMnVz9KBL13aNYa+ipWZ64FHqxuf/ytgBfBr4EHgK5n5StX1MRof+v4G+D1wdwvKlfqFo2xK76O6euevM/Pft7oWqT94pC9JBfFIX5IK4pG+JBXE0Jekghj6klQQQ1+SCmLoS1JB/j/p34tDPOnMRwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "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": [
    "# no meta\n",
    "report_results('topn', result, target_metric);"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "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></th>\n",
       "      <th>BX</th>\n",
       "      <th>ML10M</th>\n",
       "      <th>ML1M</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>type</th>\n",
       "      <th>metric</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>relevance</th>\n",
       "      <th>hr</th>\n",
       "      <td>0.050140</td>\n",
       "      <td>0.268534</td>\n",
       "      <td>0.210568</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>ranking</th>\n",
       "      <th>mrr</th>\n",
       "      <td>0.022984</td>\n",
       "      <td>0.116333</td>\n",
       "      <td>0.088191</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                        BX     ML10M      ML1M\n",
       "type      metric                              \n",
       "relevance hr      0.050140  0.268534  0.210568\n",
       "ranking   mrr     0.022984  0.116333  0.088191"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.concat({lbl: res.mean(level='top-n').loc[10, :'ranking'] for lbl, res in result.items()}, axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### saving data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "save_results(experiment_name, cv=result)"
   ]
  },
  {
   "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
}