s416082/WSS-project
1
0
Fork 0
forked from robkw/introduction_to_recommender_systems
Code Issues Pull Requests Projects Releases Wiki Activity
fb56daefbd
WSS-project/P2. Evaluation.ipynb

1693 lines
53 KiB
Plaintext
Raw Normal View History

first 2 lessons
2021-03-20 20:01:22 +01:00
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Prepare test set"
]
},
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"slideshow": {
"slide_type": "-"
}
},
"outputs": [],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"import scipy.sparse as sparse\n",
"from collections import defaultdict\n",
"from itertools import chain\n",
"import random\n",
"from tqdm import tqdm\n",
"\n",
"# In evaluation we do not load train set - it is not needed\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"test = pd.read_csv(\"./Datasets/ml-100k/test.csv\", sep=\"\\t\", header=None)\n",
"test.columns = [\"user\", \"item\", \"rating\", \"timestamp\"]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"test[\"user_code\"] = test[\"user\"].astype(\"category\").cat.codes\n",
"test[\"item_code\"] = test[\"item\"].astype(\"category\").cat.codes\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"user_code_id = dict(enumerate(test[\"user\"].astype(\"category\").cat.categories))\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"user_id_code = dict((v, k) for k, v in user_code_id.items())\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"item_code_id = dict(enumerate(test[\"item\"].astype(\"category\").cat.categories))\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"item_id_code = dict((v, k) for k, v in item_code_id.items())\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"test_ui = sparse.csr_matrix((test[\"rating\"], (test[\"user_code\"], test[\"item_code\"])))"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Estimations metrics"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"estimations_df = pd.read_csv(\n",
" \"Recommendations generated/ml-100k/Ready_Baseline_estimations.csv\", header=None\n",
")\n",
"estimations_df.columns = [\"user\", \"item\", \"score\"]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"estimations_df[\"user_code\"] = [user_id_code[user] for user in estimations_df[\"user\"]]\n",
"estimations_df[\"item_code\"] = [item_id_code[item] for item in estimations_df[\"item\"]]\n",
"estimations = sparse.csr_matrix(\n",
" (\n",
" estimations_df[\"score\"],\n",
" (estimations_df[\"user_code\"], estimations_df[\"item_code\"]),\n",
" ),\n",
" shape=test_ui.shape,\n",
")"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"def estimations_metrics(test_ui, estimations):\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" result = []\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" RMSE = (np.sum((estimations.data - test_ui.data) ** 2) / estimations.nnz) ** (1 / 2)\n",
" result.append([\"RMSE\", RMSE])\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" MAE = np.sum(abs(estimations.data - test_ui.data)) / estimations.nnz\n",
" result.append([\"MAE\", MAE])\n",
"\n",
" df_result = (pd.DataFrame(list(zip(*result))[1])).T\n",
" df_result.columns = list(zip(*result))[0]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" return df_result"
]
},
{
"cell_type": "code",
"execution_count": 4,
"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>RMSE</th>\n",
" <th>MAE</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>0.949459</td>\n",
" <td>0.752487</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" RMSE MAE\n",
"0 0.949459 0.752487"
]
},
"execution_count": 4,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# in case of error (in the laboratories) you might have to switch to the other version of pandas\n",
"# try !pip3 install pandas=='1.0.3' (or pip if you use python 2) and restart the kernel\n",
"\n",
"estimations_metrics(test_ui, estimations)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Ranking metrics"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"data": {
"text/plain": [
"array([[663, 475, 62, ..., 472, 269, 503],\n",
" [ 48, 313, 475, ..., 591, 175, 466],\n",
" [351, 313, 475, ..., 591, 175, 466],\n",
" ...,\n",
" [259, 313, 475, ..., 11, 591, 175],\n",
" [ 33, 313, 475, ..., 11, 591, 175],\n",
" [ 77, 313, 475, ..., 11, 591, 175]])"
]
},
"execution_count": 5,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import numpy as np\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"\n",
"reco = np.loadtxt(\n",
" \"Recommendations generated/ml-100k/Ready_Baseline_reco.csv\", delimiter=\",\"\n",
")\n",
"# Let's ignore scores - they are not used in evaluation:\n",
"users = reco[:, :1]\n",
"items = reco[:, 1::2]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"# Let's use inner ids instead of real ones\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"users = np.vectorize(lambda x: user_id_code.setdefault(x, -1))(users)\n",
"items = np.vectorize(lambda x: item_id_code.setdefault(x, -1))(items)\n",
"reco = np.concatenate((users, items), axis=1)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"reco"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"def ranking_metrics(test_ui, reco, super_reactions=[], topK=10):\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"\n",
" nb_items = test_ui.shape[1]\n",
" (\n",
" relevant_users,\n",
" super_relevant_users,\n",
" prec,\n",
" rec,\n",
" F_1,\n",
" F_05,\n",
" prec_super,\n",
" rec_super,\n",
" ndcg,\n",
" mAP,\n",
" MRR,\n",
" LAUC,\n",
" HR,\n",
" ) = (0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0)\n",
"\n",
" cg = 1.0 / np.log2(np.arange(2, topK + 2))\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" cg_sum = np.cumsum(cg)\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"\n",
" for (nb_user, user) in tqdm(enumerate(reco[:, 0])):\n",
" u_rated_items = test_ui.indices[test_ui.indptr[user] : test_ui.indptr[user + 1]]\n",
" nb_u_rated_items = len(u_rated_items)\n",
" if (\n",
" nb_u_rated_items > 0\n",
" ): # skip users with no items in test set (still possible that there will be no super items)\n",
" relevant_users += 1\n",
"\n",
" u_super_items = u_rated_items[\n",
" np.vectorize(lambda x: x in super_reactions)(\n",
" test_ui.data[test_ui.indptr[user] : test_ui.indptr[user + 1]]\n",
" )\n",
" ]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" # more natural seems u_super_items=[item for item in u_rated_items if test_ui[user,item] in super_reactions]\n",
" # but accesing test_ui[user,item] is expensive -we should avoid doing it\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" if len(u_super_items) > 0:\n",
" super_relevant_users += 1\n",
"\n",
" user_successes = np.zeros(topK)\n",
" nb_user_successes = 0\n",
" user_super_successes = np.zeros(topK)\n",
" nb_user_super_successes = 0\n",
"\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" # evaluation\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" for (item_position, item) in enumerate(reco[nb_user, 1 : topK + 1]):\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" if item in u_rated_items:\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" user_successes[item_position] = 1\n",
" nb_user_successes += 1\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" if item in u_super_items:\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" user_super_successes[item_position] = 1\n",
" nb_user_super_successes += 1\n",
"\n",
" prec_u = nb_user_successes / topK\n",
" prec += prec_u\n",
"\n",
" rec_u = nb_user_successes / nb_u_rated_items\n",
" rec += rec_u\n",
"\n",
" F_1 += 2 * (prec_u * rec_u) / (prec_u + rec_u) if prec_u + rec_u > 0 else 0\n",
" F_05 += (\n",
" (0.5 ** 2 + 1) * (prec_u * rec_u) / (0.5 ** 2 * prec_u + rec_u)\n",
" if prec_u + rec_u > 0\n",
" else 0\n",
" )\n",
"\n",
" prec_super += nb_user_super_successes / topK\n",
" rec_super += nb_user_super_successes / max(\n",
" len(u_super_items), 1\n",
" ) # to set 0 if no super items\n",
" ndcg += np.dot(user_successes, cg) / cg_sum[min(topK, nb_u_rated_items) - 1]\n",
"\n",
" cumsum_successes = np.cumsum(user_successes)\n",
" mAP += np.dot(\n",
" cumsum_successes / np.arange(1, topK + 1), user_successes\n",
" ) / min(topK, nb_u_rated_items)\n",
" MRR += (\n",
" 1 / (user_successes.nonzero()[0][0] + 1)\n",
" if user_successes.nonzero()[0].size > 0\n",
" else 0\n",
" )\n",
" LAUC += (\n",
" np.dot(cumsum_successes, 1 - user_successes)\n",
" + (nb_user_successes + nb_u_rated_items)\n",
" / 2\n",
" * ((nb_items - nb_u_rated_items) - (topK - nb_user_successes))\n",
" ) / ((nb_items - nb_u_rated_items) * nb_u_rated_items)\n",
"\n",
" HR += nb_user_successes > 0\n",
"\n",
" result = []\n",
" result.append((\"precision\", prec / relevant_users))\n",
" result.append((\"recall\", rec / relevant_users))\n",
" result.append((\"F_1\", F_1 / relevant_users))\n",
" result.append((\"F_05\", F_05 / relevant_users))\n",
" result.append((\"precision_super\", prec_super / super_relevant_users))\n",
" result.append((\"recall_super\", rec_super / super_relevant_users))\n",
" result.append((\"NDCG\", ndcg / relevant_users))\n",
" result.append((\"mAP\", mAP / relevant_users))\n",
" result.append((\"MRR\", MRR / relevant_users))\n",
" result.append((\"LAUC\", LAUC / relevant_users))\n",
" result.append((\"HR\", HR / relevant_users))\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" df_result = (pd.DataFrame(list(zip(*result))[1])).T\n",
" df_result.columns = list(zip(*result))[0]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" return df_result"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
2nd meeting
2021-04-16 22:41:06 +02:00
"943it [00:00, 9434.06it/s]\n"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"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>precision</th>\n",
" <th>recall</th>\n",
" <th>F_1</th>\n",
" <th>F_05</th>\n",
" <th>precision_super</th>\n",
" <th>recall_super</th>\n",
" <th>NDCG</th>\n",
" <th>mAP</th>\n",
" <th>MRR</th>\n",
" <th>LAUC</th>\n",
" <th>HR</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>0.09141</td>\n",
" <td>0.037652</td>\n",
" <td>0.04603</td>\n",
" <td>0.061286</td>\n",
" <td>0.079614</td>\n",
" <td>0.056463</td>\n",
" <td>0.095957</td>\n",
" <td>0.043178</td>\n",
" <td>0.198193</td>\n",
" <td>0.515501</td>\n",
" <td>0.437964</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" precision recall F_1 F_05 precision_super recall_super \\\n",
"0 0.09141 0.037652 0.04603 0.061286 0.079614 0.056463 \n",
"\n",
" NDCG mAP MRR LAUC HR \n",
"0 0.095957 0.043178 0.198193 0.515501 0.437964 "
]
},
"execution_count": 7,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"ranking_metrics(test_ui, reco, super_reactions=[4, 5], topK=10)"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Diversity metrics"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"def diversity_metrics(test_ui, reco, topK=10):\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"\n",
" frequencies = defaultdict(int)\n",
"\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" # let's assign 0 to all items in test set\n",
" for item in list(set(test_ui.indices)):\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" frequencies[item] = 0\n",
"\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" # counting frequencies\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" for item in reco[:, 1:].flat:\n",
" frequencies[item] += 1\n",
"\n",
" nb_reco_outside_test = frequencies[-1]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" del frequencies[-1]\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"\n",
" frequencies = np.array(list(frequencies.values()))\n",
"\n",
" nb_rec_items = len(frequencies[frequencies > 0])\n",
" nb_reco_inside_test = np.sum(frequencies)\n",
"\n",
" frequencies = frequencies / np.sum(frequencies)\n",
" frequencies = np.sort(frequencies)\n",
"\n",
" with np.errstate(\n",
" divide=\"ignore\"\n",
" ): # let's put zeros put items with 0 frequency and ignore division warning\n",
" log_frequencies = np.nan_to_num(np.log(frequencies), posinf=0, neginf=0)\n",
"\n",
" result = []\n",
" result.append(\n",
" (\n",
" \"Reco in test\",\n",
" nb_reco_inside_test / (nb_reco_inside_test + nb_reco_outside_test),\n",
" )\n",
" )\n",
" result.append((\"Test coverage\", nb_rec_items / test_ui.shape[1]))\n",
" result.append((\"Shannon\", -np.dot(frequencies, log_frequencies)))\n",
" result.append(\n",
" (\n",
" \"Gini\",\n",
" np.dot(frequencies, np.arange(1 - len(frequencies), len(frequencies), 2))\n",
" / (len(frequencies) - 1),\n",
" )\n",
" )\n",
"\n",
" df_result = (pd.DataFrame(list(zip(*result))[1])).T\n",
" df_result.columns = list(zip(*result))[0]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" return df_result"
]
},
{
"cell_type": "code",
"execution_count": 9,
"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>Reco in test</th>\n",
" <th>Test coverage</th>\n",
" <th>Shannon</th>\n",
" <th>Gini</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>1.0</td>\n",
" <td>0.033911</td>\n",
" <td>2.836513</td>\n",
" <td>0.991139</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Reco in test Test coverage Shannon Gini\n",
"0 1.0 0.033911 2.836513 0.991139"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"# in case of errors try !pip3 install numpy==1.18.4 (or pip if you use python 2) and restart the kernel\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"x = diversity_metrics(test_ui, reco, topK=10)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"x"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# To be used in other notebooks"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
2nd meeting
2021-04-16 22:41:06 +02:00
"943it [00:00, 11012.47it/s]\n"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"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>RMSE</th>\n",
" <th>MAE</th>\n",
" <th>precision</th>\n",
" <th>recall</th>\n",
" <th>F_1</th>\n",
" <th>F_05</th>\n",
" <th>precision_super</th>\n",
" <th>recall_super</th>\n",
" <th>NDCG</th>\n",
" <th>mAP</th>\n",
" <th>MRR</th>\n",
" <th>LAUC</th>\n",
" <th>HR</th>\n",
" <th>Reco in test</th>\n",
" <th>Test coverage</th>\n",
" <th>Shannon</th>\n",
" <th>Gini</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>0.949459</td>\n",
" <td>0.752487</td>\n",
" <td>0.09141</td>\n",
" <td>0.037652</td>\n",
" <td>0.04603</td>\n",
" <td>0.061286</td>\n",
" <td>0.079614</td>\n",
" <td>0.056463</td>\n",
" <td>0.095957</td>\n",
" <td>0.043178</td>\n",
" <td>0.198193</td>\n",
" <td>0.515501</td>\n",
" <td>0.437964</td>\n",
" <td>1.0</td>\n",
" <td>0.033911</td>\n",
" <td>2.836513</td>\n",
" <td>0.991139</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" RMSE MAE precision recall F_1 F_05 \\\n",
"0 0.949459 0.752487 0.09141 0.037652 0.04603 0.061286 \n",
"\n",
" precision_super recall_super NDCG mAP MRR LAUC \\\n",
"0 0.079614 0.056463 0.095957 0.043178 0.198193 0.515501 \n",
"\n",
" HR Reco in test Test coverage Shannon Gini \n",
"0 0.437964 1.0 0.033911 2.836513 0.991139 "
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"import evaluation_measures as ev\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"estimations_df = pd.read_csv(\n",
" \"Recommendations generated/ml-100k/Ready_Baseline_estimations.csv\", header=None\n",
")\n",
"reco = np.loadtxt(\n",
" \"Recommendations generated/ml-100k/Ready_Baseline_reco.csv\", delimiter=\",\"\n",
")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"ev.evaluate(\n",
" test=pd.read_csv(\"./Datasets/ml-100k/test.csv\", sep=\"\\t\", header=None),\n",
" estimations_df=estimations_df,\n",
" reco=reco,\n",
" super_reactions=[4, 5],\n",
")\n",
"# also you can just type ev.evaluate_all(estimations_df, reco) - I put above values as default"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
2nd meeting
2021-04-16 22:41:06 +02:00
"943it [00:00, 10346.82it/s]\n",
"943it [00:00, 11772.32it/s]\n",
"943it [00:00, 10636.62it/s]\n",
"943it [00:00, 10767.92it/s]\n",
"943it [00:00, 12019.93it/s]\n"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
}
],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"dir_path = \"Recommendations generated/ml-100k/\"\n",
"super_reactions = [4, 5]\n",
"test = pd.read_csv(\"./Datasets/ml-100k/test.csv\", sep=\"\\t\", header=None)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"df = ev.evaluate_all(test, dir_path, super_reactions)\n",
"# also you can just type ev.evaluate_all() - I put above values as default"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "code",
"execution_count": 12,
"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>Model</th>\n",
" <th>RMSE</th>\n",
" <th>MAE</th>\n",
" <th>precision</th>\n",
" <th>recall</th>\n",
" <th>F_1</th>\n",
" <th>F_05</th>\n",
" <th>precision_super</th>\n",
" <th>recall_super</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Self_TopPop</td>\n",
" <td>2.508258</td>\n",
" <td>2.217909</td>\n",
" <td>0.188865</td>\n",
" <td>0.116919</td>\n",
" <td>0.118732</td>\n",
" <td>0.141584</td>\n",
" <td>0.130472</td>\n",
" <td>0.137473</td>\n",
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Ready_Baseline</td>\n",
" <td>0.949459</td>\n",
" <td>0.752487</td>\n",
" <td>0.091410</td>\n",
" <td>0.037652</td>\n",
" <td>0.046030</td>\n",
" <td>0.061286</td>\n",
" <td>0.079614</td>\n",
" <td>0.056463</td>\n",
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>Ready_Random</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>1.521845</td>\n",
" <td>1.225949</td>\n",
" <td>0.047190</td>\n",
" <td>0.020753</td>\n",
" <td>0.024810</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>0.032269</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>0.029506</td>\n",
" <td>0.023707</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>Self_TopRated</td>\n",
" <td>1.030712</td>\n",
" <td>0.820904</td>\n",
" <td>0.000954</td>\n",
" <td>0.000188</td>\n",
" <td>0.000298</td>\n",
" <td>0.000481</td>\n",
" <td>0.000644</td>\n",
" <td>0.000223</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Self_BaselineUI</td>\n",
" <td>0.967585</td>\n",
" <td>0.762740</td>\n",
" <td>0.000954</td>\n",
" <td>0.000170</td>\n",
" <td>0.000278</td>\n",
" <td>0.000463</td>\n",
" <td>0.000644</td>\n",
" <td>0.000189</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
slides updated
2021-03-20 20:13:28 +01:00
" Model RMSE MAE precision recall F_1 \\\n",
"0 Self_TopPop 2.508258 2.217909 0.188865 0.116919 0.118732 \n",
"0 Ready_Baseline 0.949459 0.752487 0.091410 0.037652 0.046030 \n",
2nd meeting
2021-04-16 22:41:06 +02:00
"0 Ready_Random 1.521845 1.225949 0.047190 0.020753 0.024810 \n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
"0 Self_TopRated 1.030712 0.820904 0.000954 0.000188 0.000298 \n",
slides updated
2021-03-20 20:13:28 +01:00
"0 Self_BaselineUI 0.967585 0.762740 0.000954 0.000170 0.000278 \n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
slides updated
2021-03-20 20:13:28 +01:00
" F_05 precision_super recall_super \n",
"0 0.141584 0.130472 0.137473 \n",
"0 0.061286 0.079614 0.056463 \n",
2nd meeting
2021-04-16 22:41:06 +02:00
"0 0.032269 0.029506 0.023707 \n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
"0 0.000481 0.000644 0.000223 \n",
slides updated
2021-03-20 20:13:28 +01:00
"0 0.000463 0.000644 0.000189 "
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
"execution_count": 12,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"df.iloc[:, :9]"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "code",
"execution_count": 13,
"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>Model</th>\n",
" <th>NDCG</th>\n",
" <th>mAP</th>\n",
" <th>MRR</th>\n",
" <th>LAUC</th>\n",
" <th>HR</th>\n",
" <th>Reco in test</th>\n",
" <th>Test coverage</th>\n",
" <th>Shannon</th>\n",
" <th>Gini</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Self_TopPop</td>\n",
" <td>0.214651</td>\n",
" <td>0.111707</td>\n",
" <td>0.400939</td>\n",
" <td>0.555546</td>\n",
" <td>0.765642</td>\n",
" <td>1.000000</td>\n",
" <td>0.038961</td>\n",
" <td>3.159079</td>\n",
" <td>0.987317</td>\n",
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Ready_Baseline</td>\n",
" <td>0.095957</td>\n",
" <td>0.043178</td>\n",
" <td>0.198193</td>\n",
" <td>0.515501</td>\n",
" <td>0.437964</td>\n",
" <td>1.000000</td>\n",
" <td>0.033911</td>\n",
" <td>2.836513</td>\n",
" <td>0.991139</td>\n",
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>Ready_Random</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>0.050075</td>\n",
" <td>0.018728</td>\n",
" <td>0.121957</td>\n",
" <td>0.506893</td>\n",
" <td>0.329799</td>\n",
" <td>0.986532</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>0.184704</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>5.099706</td>\n",
" <td>0.907217</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>Self_TopRated</td>\n",
" <td>0.001043</td>\n",
" <td>0.000335</td>\n",
" <td>0.003348</td>\n",
" <td>0.496433</td>\n",
" <td>0.009544</td>\n",
" <td>0.699046</td>\n",
" <td>0.005051</td>\n",
" <td>1.945910</td>\n",
" <td>0.995669</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Self_BaselineUI</td>\n",
" <td>0.000752</td>\n",
" <td>0.000168</td>\n",
" <td>0.001677</td>\n",
" <td>0.496424</td>\n",
" <td>0.009544</td>\n",
" <td>0.600530</td>\n",
" <td>0.005051</td>\n",
" <td>1.803126</td>\n",
" <td>0.996380</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
slides updated
2021-03-20 20:13:28 +01:00
" Model NDCG mAP MRR LAUC HR \\\n",
"0 Self_TopPop 0.214651 0.111707 0.400939 0.555546 0.765642 \n",
"0 Ready_Baseline 0.095957 0.043178 0.198193 0.515501 0.437964 \n",
2nd meeting
2021-04-16 22:41:06 +02:00
"0 Ready_Random 0.050075 0.018728 0.121957 0.506893 0.329799 \n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
"0 Self_TopRated 0.001043 0.000335 0.003348 0.496433 0.009544 \n",
slides updated
2021-03-20 20:13:28 +01:00
"0 Self_BaselineUI 0.000752 0.000168 0.001677 0.496424 0.009544 \n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
" Reco in test Test coverage Shannon Gini \n",
"0 1.000000 0.038961 3.159079 0.987317 \n",
"0 1.000000 0.033911 2.836513 0.991139 \n",
2nd meeting
2021-04-16 22:41:06 +02:00
"0 0.986532 0.184704 5.099706 0.907217 \n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
"0 0.699046 0.005051 1.945910 0.995669 \n",
slides updated
2021-03-20 20:13:28 +01:00
"0 0.600530 0.005051 1.803126 0.996380 "
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
"execution_count": 13,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"df.iloc[:, np.append(0, np.arange(9, df.shape[1]))]"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Check metrics on toy dataset"
]
},
{
"cell_type": "code",
"execution_count": 14,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
2nd meeting
2021-04-16 22:41:06 +02:00
"3it [00:00, 5771.98it/s]\n"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"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>Model</th>\n",
" <th>RMSE</th>\n",
" <th>MAE</th>\n",
" <th>precision</th>\n",
" <th>recall</th>\n",
" <th>F_1</th>\n",
" <th>F_05</th>\n",
" <th>precision_super</th>\n",
" <th>recall_super</th>\n",
" <th>NDCG</th>\n",
" <th>mAP</th>\n",
" <th>MRR</th>\n",
" <th>LAUC</th>\n",
" <th>HR</th>\n",
" <th>Reco in test</th>\n",
" <th>Test coverage</th>\n",
" <th>Shannon</th>\n",
" <th>Gini</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>Self_BaselineUI</td>\n",
" <td>1.612452</td>\n",
" <td>1.4</td>\n",
" <td>0.444444</td>\n",
" <td>0.888889</td>\n",
" <td>0.555556</td>\n",
" <td>0.478632</td>\n",
" <td>0.333333</td>\n",
" <td>0.75</td>\n",
" <td>0.676907</td>\n",
" <td>0.574074</td>\n",
" <td>0.611111</td>\n",
" <td>0.638889</td>\n",
" <td>1.0</td>\n",
" <td>0.888889</td>\n",
" <td>0.8</td>\n",
" <td>1.386294</td>\n",
" <td>0.25</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" Model RMSE MAE precision recall F_1 F_05 \\\n",
"0 Self_BaselineUI 1.612452 1.4 0.444444 0.888889 0.555556 0.478632 \n",
"\n",
" precision_super recall_super NDCG mAP MRR LAUC HR \\\n",
"0 0.333333 0.75 0.676907 0.574074 0.611111 0.638889 1.0 \n",
"\n",
" Reco in test Test coverage Shannon Gini \n",
"0 0.888889 0.8 1.386294 0.25 "
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Training data:\n"
]
},
{
"data": {
"text/plain": [
"matrix([[3, 4, 0, 0, 5, 0, 0, 4],\n",
" [0, 1, 2, 3, 0, 0, 0, 0],\n",
" [0, 0, 0, 5, 0, 3, 4, 0]])"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Test data:\n"
]
},
{
"data": {
"text/plain": [
"matrix([[0, 0, 0, 0, 0, 0, 3, 0],\n",
" [0, 0, 0, 0, 5, 0, 0, 0],\n",
" [5, 0, 4, 0, 0, 0, 0, 2]])"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Recommendations:\n"
]
},
{
"data": {
"text/html": [
"<div>\n",
"<style scoped>\n",
" .dataframe tbody tr th:only-of-type {\n",
" vertical-align: middle;\n",
" }\n",
"\n",
" .dataframe tbody tr th {\n",
" vertical-align: top;\n",
" }\n",
"\n",
" .dataframe thead th {\n",
" text-align: right;\n",
" }\n",
"</style>\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>0</th>\n",
" <th>1</th>\n",
" <th>2</th>\n",
" <th>3</th>\n",
" <th>4</th>\n",
" <th>5</th>\n",
" <th>6</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>0</td>\n",
" <td>30</td>\n",
" <td>5.0</td>\n",
" <td>20</td>\n",
" <td>4.0</td>\n",
" <td>60</td>\n",
" <td>4.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>10</td>\n",
" <td>40</td>\n",
" <td>3.0</td>\n",
" <td>60</td>\n",
" <td>2.0</td>\n",
" <td>70</td>\n",
" <td>2.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>20</td>\n",
" <td>40</td>\n",
" <td>5.0</td>\n",
" <td>20</td>\n",
" <td>4.0</td>\n",
" <td>70</td>\n",
" <td>4.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" 0 1 2 3 4 5 6\n",
"0 0 30 5.0 20 4.0 60 4.0\n",
"1 10 40 3.0 60 2.0 70 2.0\n",
"2 20 40 5.0 20 4.0 70 4.0"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Estimations:\n"
]
},
{
"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>user</th>\n",
" <th>item</th>\n",
" <th>est_score</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td>0</td>\n",
" <td>60</td>\n",
" <td>4.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td>10</td>\n",
" <td>40</td>\n",
" <td>3.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td>20</td>\n",
" <td>0</td>\n",
" <td>3.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td>20</td>\n",
" <td>20</td>\n",
" <td>4.0</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td>20</td>\n",
" <td>70</td>\n",
" <td>4.0</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" user item est_score\n",
"0 0 60 4.0\n",
"1 10 40 3.0\n",
"2 20 0 3.0\n",
"3 20 20 4.0\n",
"4 20 70 4.0"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"import helpers\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"dir_path = \"Recommendations generated/toy-example/\"\n",
"super_reactions = [4, 5]\n",
"test = pd.read_csv(\"./Datasets/toy-example/test.csv\", sep=\"\\t\", header=None)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
"display(ev.evaluate_all(test, dir_path, super_reactions, topK=3))\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"# also you can just type ev.evaluate_all() - I put above values as default\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"toy_train_read = pd.read_csv(\n",
" \"./Datasets/toy-example/train.csv\",\n",
" sep=\"\\t\",\n",
" header=None,\n",
" names=[\"user\", \"item\", \"rating\", \"timestamp\"],\n",
")\n",
"toy_test_read = pd.read_csv(\n",
" \"./Datasets/toy-example/test.csv\",\n",
" sep=\"\\t\",\n",
" header=None,\n",
" names=[\"user\", \"item\", \"rating\", \"timestamp\"],\n",
")\n",
"reco = pd.read_csv(\n",
" \"Recommendations generated/toy-example/Self_BaselineUI_reco.csv\", header=None\n",
")\n",
"estimations = pd.read_csv(\n",
" \"Recommendations generated/toy-example/Self_BaselineUI_estimations.csv\",\n",
" names=[\"user\", \"item\", \"est_score\"],\n",
")\n",
"(\n",
" toy_train_ui,\n",
" toy_test_ui,\n",
" toy_user_code_id,\n",
" toy_user_id_code,\n",
" toy_item_code_id,\n",
" toy_item_id_code,\n",
") = helpers.data_to_csr(toy_train_read, toy_test_read)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Training data:\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"display(toy_train_ui.todense())\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Test data:\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"display(toy_test_ui.todense())\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Recommendations:\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"display(reco)\n",
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Estimations:\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"display(estimations)"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# Sample recommendations"
]
},
{
"cell_type": "code",
slides updated
2021-03-20 20:13:28 +01:00
"execution_count": 15,
first 2 lessons
2021-03-20 20:01:22 +01:00
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Here is what user rated high:\n"
]
},
{
"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>user</th>\n",
" <th>rating</th>\n",
" <th>title</th>\n",
" <th>genres</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>57482</th>\n",
" <td>2</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Emma (1996)</td>\n",
" <td>Drama, Romance</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>54506</th>\n",
" <td>2</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Sense and Sensibility (1995)</td>\n",
" <td>Drama, Romance</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>40581</th>\n",
" <td>2</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Titanic (1997)</td>\n",
" <td>Action, Drama, Romance</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>2949</th>\n",
" <td>2</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Star Wars (1977)</td>\n",
" <td>Action, Adventure, Romance, Sci-Fi, War</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>69653</th>\n",
" <td>2</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Wings of the Dove, The (1997)</td>\n",
" <td>Drama, Romance, Thriller</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>7906</th>\n",
" <td>2</td>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>As Good As It Gets (1997)</td>\n",
" <td>Comedy, Drama</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>69400</th>\n",
" <td>2</td>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Shall We Dance? (1996)</td>\n",
" <td>Comedy</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>14469</th>\n",
" <td>2</td>\n",
finished first 2 lectures
2021-03-23 21:52:46 +01:00
" <td>5</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Fargo (1996)</td>\n",
" <td>Crime, Drama, Thriller</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>46151</th>\n",
" <td>2</td>\n",
" <td>5</td>\n",
" <td>L.A. Confidential (1997)</td>\n",
" <td>Crime, Film-Noir, Mystery, Thriller</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>67293</th>\n",
" <td>2</td>\n",
" <td>5</td>\n",
" <td>Good Will Hunting (1997)</td>\n",
" <td>Drama</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>20923</th>\n",
" <td>2</td>\n",
" <td>5</td>\n",
" <td>Secrets &amp; Lies (1996)</td>\n",
" <td>Drama</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>52921</th>\n",
" <td>2</td>\n",
" <td>5</td>\n",
" <td>Kolya (1996)</td>\n",
" <td>Comedy</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>50103</th>\n",
" <td>2</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>4</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Mrs. Brown (Her Majesty, Mrs. Brown) (1997)</td>\n",
" <td>Drama, Romance</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>51972</th>\n",
" <td>2</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>4</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Mighty Aphrodite (1995)</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>Comedy</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>515</th>\n",
" <td>2</td>\n",
last changes
2021-03-26 21:00:52 +01:00
" <td>4</td>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <td>Heat (1995)</td>\n",
" <td>Action, Crime, Thriller</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
2nd meeting
2021-04-16 22:41:06 +02:00
" user rating title \\\n",
"57482 2 5 Emma (1996) \n",
"54506 2 5 Sense and Sensibility (1995) \n",
"40581 2 5 Titanic (1997) \n",
"2949 2 5 Star Wars (1977) \n",
"69653 2 5 Wings of the Dove, The (1997) \n",
"7906 2 5 As Good As It Gets (1997) \n",
"69400 2 5 Shall We Dance? (1996) \n",
"14469 2 5 Fargo (1996) \n",
"46151 2 5 L.A. Confidential (1997) \n",
"67293 2 5 Good Will Hunting (1997) \n",
"20923 2 5 Secrets & Lies (1996) \n",
"52921 2 5 Kolya (1996) \n",
"50103 2 4 Mrs. Brown (Her Majesty, Mrs. Brown) (1997) \n",
"51972 2 4 Mighty Aphrodite (1995) \n",
"515 2 4 Heat (1995) \n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" genres \n",
"57482 Drama, Romance \n",
"54506 Drama, Romance \n",
"40581 Action, Drama, Romance \n",
"2949 Action, Adventure, Romance, Sci-Fi, War \n",
"69653 Drama, Romance, Thriller \n",
"7906 Comedy, Drama \n",
"69400 Comedy \n",
"14469 Crime, Drama, Thriller \n",
"46151 Crime, Film-Noir, Mystery, Thriller \n",
"67293 Drama \n",
"20923 Drama \n",
"52921 Comedy \n",
"50103 Drama, Romance \n",
"51972 Comedy \n",
"515 Action, Crime, Thriller "
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stdout",
"output_type": "stream",
"text": [
"Here is what we recommend:\n"
]
},
{
"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>user</th>\n",
" <th>rec_nb</th>\n",
" <th>title</th>\n",
" <th>genres</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>1</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>1</td>\n",
" <td>Great Day in Harlem, A (1994)</td>\n",
" <td>Documentary</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>943</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>2</td>\n",
" <td>Tough and Deadly (1995)</td>\n",
" <td>Action, Drama, Thriller</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>1885</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>3</td>\n",
" <td>Aiqing wansui (1994)</td>\n",
" <td>Drama</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>2827</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>4</td>\n",
" <td>Delta of Venus (1994)</td>\n",
" <td>Drama</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>3769</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>5</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>Someone Else's America (1995)</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>Drama</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>4711</th>\n",
" <td>2.0</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>6</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>Saint of Fort Washington, The (1993)</td>\n",
" <td>Drama</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>5653</th>\n",
" <td>2.0</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>7</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>Celestial Clockwork (1994)</td>\n",
" <td>Comedy</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>6595</th>\n",
" <td>2.0</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>8</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>Some Mother's Son (1996)</td>\n",
" <td>Drama</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>8489</th>\n",
" <td>2.0</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>9</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>Maya Lin: A Strong Clear Vision (1994)</td>\n",
" <td>Documentary</td>\n",
" </tr>\n",
" <tr>\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" <th>7536</th>\n",
" <td>2.0</td>\n",
slides updated
2021-03-20 20:13:28 +01:00
" <td>10</td>\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
" <td>Prefontaine (1997)</td>\n",
" <td>Drama</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
2nd meeting
2021-04-16 22:41:06 +02:00
" user rec_nb title \\\n",
"1 2.0 1 Great Day in Harlem, A (1994) \n",
"943 2.0 2 Tough and Deadly (1995) \n",
"1885 2.0 3 Aiqing wansui (1994) \n",
"2827 2.0 4 Delta of Venus (1994) \n",
"3769 2.0 5 Someone Else's America (1995) \n",
"4711 2.0 6 Saint of Fort Washington, The (1993) \n",
"5653 2.0 7 Celestial Clockwork (1994) \n",
"6595 2.0 8 Some Mother's Son (1996) \n",
"8489 2.0 9 Maya Lin: A Strong Clear Vision (1994) \n",
"7536 2.0 10 Prefontaine (1997) \n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
" genres \n",
2nd meeting
2021-04-16 22:41:06 +02:00
"1 Documentary \n",
"943 Action, Drama, Thriller \n",
"1885 Drama \n",
"2827 Drama \n",
"3769 Drama \n",
"4711 Drama \n",
"5653 Comedy \n",
"6595 Drama \n",
"8489 Documentary \n",
"7536 Drama "
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
slides updated
2021-03-20 20:13:28 +01:00
"execution_count": 15,
first 2 lessons
2021-03-20 20:01:22 +01:00
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"train = pd.read_csv(\n",
" \"./Datasets/ml-100k/train.csv\",\n",
" sep=\"\\t\",\n",
" header=None,\n",
" names=[\"user\", \"item\", \"rating\", \"timestamp\"],\n",
")\n",
"items = pd.read_csv(\"./Datasets/ml-100k/movies.csv\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"user = random.choice(list(set(train[\"user\"])))\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"train_content = pd.merge(train, items, left_on=\"item\", right_on=\"id\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Here is what user rated high:\")\n",
"display(\n",
" train_content[train_content[\"user\"] == user][\n",
" [\"user\", \"rating\", \"title\", \"genres\"]\n",
" ].sort_values(by=\"rating\", ascending=False)[:15]\n",
")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"reco = np.loadtxt(\n",
" \"Recommendations generated/ml-100k/Self_BaselineUI_reco.csv\", delimiter=\",\"\n",
")\n",
"items = pd.read_csv(\"./Datasets/ml-100k/movies.csv\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"# Let's ignore scores - they are not used in evaluation:\n",
"reco_users = reco[:, :1]\n",
"reco_items = reco[:, 1::2]\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"# Let's put them into one array\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"reco = np.concatenate((reco_users, reco_items), axis=1)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
"# Let's rebuild it user-item dataframe\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"recommended = []\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"for row in reco:\n",
" for rec_nb, entry in enumerate(row[1:]):\n",
2nd meeting
2021-04-16 22:41:06 +02:00
" recommended.append((row[0], rec_nb + 1, entry))\n",
"recommended = pd.DataFrame(recommended, columns=[\"user\", \"rec_nb\", \"item\"])\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"recommended_content = pd.merge(recommended, items, left_on=\"item\", right_on=\"id\")\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"\n",
2nd meeting
2021-04-16 22:41:06 +02:00
"print(\"Here is what we recommend:\")\n",
"recommended_content[recommended_content[\"user\"] == user][\n",
" [\"user\", \"rec_nb\", \"title\", \"genres\"]\n",
"].sort_values(by=\"rec_nb\")"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
last changes
2021-03-26 21:00:52 +01:00
"# project task 2: implement some other evaluation measure"
first 2 lessons
2021-03-20 20:01:22 +01:00
]
},
{
"cell_type": "code",
slides updated
2021-03-20 20:13:28 +01:00
"execution_count": 16,
first 2 lessons
2021-03-20 20:01:22 +01:00
"metadata": {},
"outputs": [],
"source": [
2nd meeting
2021-04-16 22:41:06 +02:00
"# it may be your idea, modification of what we have already implemented\n",
"# (for example Hit2 rate which would count as a success users whoreceived at least 2 relevant recommendations)\n",
first 2 lessons
2021-03-20 20:01:22 +01:00
"# or something well-known\n",
"# expected output: modification of evaluation_measures.py such that evaluate_all will also display your measure"
]
}
],
"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",
2nd meeting
2021-04-16 22:41:06 +02:00
"version": "3.8.5"
first 2 lessons
2021-03-20 20:01:22 +01:00
}
},
"nbformat": 4,
"nbformat_minor": 4
}
Reference in New Issue Copy Permalink