knn_ranking_task2.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas\n",
    "import numpy as np\n",
    "import sklearn as sk\n",
    "\n",
    "from sklearn.preprocessing import OneHotEncoder\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "import time\n",
    "import pickle\n",
    "\n",
    "import time\n",
    "%matplotlib inline\n",
    "from random import randint\n",
    "\n",
    "from scipy.spatial import KDTree"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>id3</th>\n",
       "      <th>user_id</th>\n",
       "      <th>date</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>714</td>\n",
       "      <td>464300</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>590</td>\n",
       "      <td>1029729</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>703</td>\n",
       "      <td>861305</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>803</td>\n",
       "      <td>733016</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>219</td>\n",
       "      <td>448027</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   id3  user_id  date\n",
       "0  714   464300     1\n",
       "1  590  1029729     1\n",
       "2  703   861305     1\n",
       "3  803   733016     1\n",
       "4  219   448027     1"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset_name = \"train.csv.zip\"\n",
    "data = pandas.read_csv(dataset_name).drop([\"id1\", \"id2\"], axis=1)\n",
    "data.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data = data[data[\"user_id\"].isin(data[\"user_id\"].unique()[:5000//5*10])][[\"user_id\", \"id3\"]].get_values()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "users = np.unique(data[:,0])\n",
    "users_vectors = dict([(user, np.zeros(931)) for user in users])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "def ohe(id3):\n",
    "    vector = np.zeros(931)\n",
    "    vector[id3] = 1\n",
    "    \n",
    "    return vector"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "users = np.unique(data[:,0])\n",
    "users_vectors = dict([(user, np.zeros(931)) for user in users])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "for row in data:\n",
    "    users_vectors[row[0]] += ohe(row[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "for user in users:\n",
    "    vec = users_vectors[user]\n",
    "    users_vectors[user] = vec/vec.sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "usvecs = []\n",
    "for key, item in users_vectors.items():\n",
    "    usvecs.append(item)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "items = [i for i in range(931)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "knn = KNeighborsRegressor(n_neighbors=150)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "KNeighborsRegressor(algorithm='auto', leaf_size=30, metric='minkowski',\n",
       "          metric_params=None, n_jobs=1, n_neighbors=30, p=2,\n",
       "          weights='uniform')"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "knn.fit(usvecs, usvecs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "items = [i for i in range(931)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "recomendations = []\n",
    "for uid in users[:5000]:\n",
    "    mean_vector = []\n",
    "    \n",
    "    vec = knn.predict([users_vectors[uid]])[0]\n",
    "    vec[users_vectors[uid] > 0] = 0\n",
    "    row = [uid] + [i[1] for i in sorted(list(zip(vec.tolist(), items)), reverse=True)[:5]]\n",
    "    \n",
    "    recomendations.append(row)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "submit = pandas.DataFrame(data=recomendations, columns=[\"user_id\", \"id3_1\", \"id3_2\", \"id3_3\", \"id3_4\", \"id3_5\"])\n",
    "submit.to_csv(\"submission.csv\", index=False)"
   ]
  },
  {
   "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": "3.6.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}