AV_Recommendation_Engine.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "problem_data=pd.read_csv(\"problem_data.csv\")\n",
    "train=pd.read_csv(\"train_submissions.csv\")\n",
    "user_data=pd.read_csv(\"user_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "test=pd.read_csv(\"test_submissions.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#merging the data from different datasheets\n",
    "train=pd.merge(train,problem_data[['problem_id','level_type','tags']],on='problem_id',how='left')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "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>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>level_type</th>\n",
       "      <th>tags</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>B</td>\n",
       "      <td>strings</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_3568</td>\n",
       "      <td>prob_2994</td>\n",
       "      <td>3</td>\n",
       "      <td>C</td>\n",
       "      <td>brute force,dp</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_1600</td>\n",
       "      <td>prob_5071</td>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>implementation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_2256</td>\n",
       "      <td>prob_703</td>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>brute force,implementation</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_2321</td>\n",
       "      <td>prob_356</td>\n",
       "      <td>1</td>\n",
       "      <td>D</td>\n",
       "      <td>dfs and similar,trees</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     user_id problem_id  attempts_range level_type                        tags\n",
       "0   user_232  prob_6507               1          B                     strings\n",
       "1  user_3568  prob_2994               3          C              brute force,dp\n",
       "2  user_1600  prob_5071               1          A              implementation\n",
       "3  user_2256   prob_703               1          A  brute force,implementation\n",
       "4  user_2321   prob_356               1          D       dfs and similar,trees"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "train=pd.merge(train,user_data[['user_id','max_rating','rank']],on='user_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>level_type</th>\n",
       "      <th>tags</th>\n",
       "      <th>max_rating</th>\n",
       "      <th>rank</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>B</td>\n",
       "      <td>strings</td>\n",
       "      <td>307.913</td>\n",
       "      <td>beginner</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_5781</td>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>implementation,math</td>\n",
       "      <td>307.913</td>\n",
       "      <td>beginner</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_137</td>\n",
       "      <td>2</td>\n",
       "      <td>B</td>\n",
       "      <td>implementation</td>\n",
       "      <td>307.913</td>\n",
       "      <td>beginner</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1550</td>\n",
       "      <td>2</td>\n",
       "      <td>A</td>\n",
       "      <td>brute force,dfs and similar</td>\n",
       "      <td>307.913</td>\n",
       "      <td>beginner</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1322</td>\n",
       "      <td>3</td>\n",
       "      <td>B</td>\n",
       "      <td>dfs and similar,shortest paths</td>\n",
       "      <td>307.913</td>\n",
       "      <td>beginner</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id problem_id  attempts_range level_type  \\\n",
       "0  user_232  prob_6507               1          B   \n",
       "1  user_232  prob_5781               1          A   \n",
       "2  user_232   prob_137               2          B   \n",
       "3  user_232  prob_1550               2          A   \n",
       "4  user_232  prob_1322               3          B   \n",
       "\n",
       "                             tags  max_rating      rank  \n",
       "0                         strings     307.913  beginner  \n",
       "1             implementation,math     307.913  beginner  \n",
       "2                  implementation     307.913  beginner  \n",
       "3     brute force,dfs and similar     307.913  beginner  \n",
       "4  dfs and similar,shortest paths     307.913  beginner  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>level_type</th>\n",
       "      <th>tags</th>\n",
       "      <th>max_rating</th>\n",
       "      <th>Rank_advanced</th>\n",
       "      <th>Rank_beginner</th>\n",
       "      <th>Rank_expert</th>\n",
       "      <th>Rank_intermediate</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>B</td>\n",
       "      <td>strings</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_5781</td>\n",
       "      <td>1</td>\n",
       "      <td>A</td>\n",
       "      <td>implementation,math</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_137</td>\n",
       "      <td>2</td>\n",
       "      <td>B</td>\n",
       "      <td>implementation</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1550</td>\n",
       "      <td>2</td>\n",
       "      <td>A</td>\n",
       "      <td>brute force,dfs and similar</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1322</td>\n",
       "      <td>3</td>\n",
       "      <td>B</td>\n",
       "      <td>dfs and similar,shortest paths</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id problem_id  attempts_range level_type  \\\n",
       "0  user_232  prob_6507               1          B   \n",
       "1  user_232  prob_5781               1          A   \n",
       "2  user_232   prob_137               2          B   \n",
       "3  user_232  prob_1550               2          A   \n",
       "4  user_232  prob_1322               3          B   \n",
       "\n",
       "                             tags  max_rating  Rank_advanced  Rank_beginner  \\\n",
       "0                         strings     307.913              0              1   \n",
       "1             implementation,math     307.913              0              1   \n",
       "2                  implementation     307.913              0              1   \n",
       "3     brute force,dfs and similar     307.913              0              1   \n",
       "4  dfs and similar,shortest paths     307.913              0              1   \n",
       "\n",
       "   Rank_expert  Rank_intermediate  \n",
       "0            0                  0  \n",
       "1            0                  0  \n",
       "2            0                  0  \n",
       "3            0                  0  \n",
       "4            0                  0  "
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#splitting the data in ranks column\n",
    "rank_dummies = pd.get_dummies(train['rank'], prefix='Rank')\n",
    "train=pd.concat([train,rank_dummies],axis=1)\n",
    "train=train.drop(['rank'],axis=1)\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "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>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>tags</th>\n",
       "      <th>max_rating</th>\n",
       "      <th>Rank_advanced</th>\n",
       "      <th>Rank_beginner</th>\n",
       "      <th>Rank_expert</th>\n",
       "      <th>Rank_intermediate</th>\n",
       "      <th>Level_A</th>\n",
       "      <th>...</th>\n",
       "      <th>Level_E</th>\n",
       "      <th>Level_F</th>\n",
       "      <th>Level_G</th>\n",
       "      <th>Level_H</th>\n",
       "      <th>Level_I</th>\n",
       "      <th>Level_J</th>\n",
       "      <th>Level_K</th>\n",
       "      <th>Level_L</th>\n",
       "      <th>Level_M</th>\n",
       "      <th>Level_N</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>strings</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_5781</td>\n",
       "      <td>1</td>\n",
       "      <td>implementation,math</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_137</td>\n",
       "      <td>2</td>\n",
       "      <td>implementation</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1550</td>\n",
       "      <td>2</td>\n",
       "      <td>brute force,dfs and similar</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1322</td>\n",
       "      <td>3</td>\n",
       "      <td>dfs and similar,shortest paths</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 23 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id problem_id  attempts_range                            tags  \\\n",
       "0  user_232  prob_6507               1                         strings   \n",
       "1  user_232  prob_5781               1             implementation,math   \n",
       "2  user_232   prob_137               2                  implementation   \n",
       "3  user_232  prob_1550               2     brute force,dfs and similar   \n",
       "4  user_232  prob_1322               3  dfs and similar,shortest paths   \n",
       "\n",
       "   max_rating  Rank_advanced  Rank_beginner  Rank_expert  Rank_intermediate  \\\n",
       "0     307.913              0              1            0                  0   \n",
       "1     307.913              0              1            0                  0   \n",
       "2     307.913              0              1            0                  0   \n",
       "3     307.913              0              1            0                  0   \n",
       "4     307.913              0              1            0                  0   \n",
       "\n",
       "   Level_A   ...     Level_E  Level_F  Level_G  Level_H  Level_I  Level_J  \\\n",
       "0        0   ...           0        0        0        0        0        0   \n",
       "1        1   ...           0        0        0        0        0        0   \n",
       "2        0   ...           0        0        0        0        0        0   \n",
       "3        1   ...           0        0        0        0        0        0   \n",
       "4        0   ...           0        0        0        0        0        0   \n",
       "\n",
       "   Level_K  Level_L  Level_M  Level_N  \n",
       "0        0        0        0        0  \n",
       "1        0        0        0        0  \n",
       "2        0        0        0        0  \n",
       "3        0        0        0        0  \n",
       "4        0        0        0        0  \n",
       "\n",
       "[5 rows x 23 columns]"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#splitting the data in level column\n",
    "level_dummies = pd.get_dummies(train['level_type'], prefix='Level')\n",
    "train=pd.concat([train,level_dummies],axis=1)\n",
    "train=train.drop(['level_type'],axis=1)\n",
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "#splitting the data in tags column\n",
    "train=pd.concat([train,train['tags'].str.get_dummies(sep=',')],axis=1)"
   ]
  },
  {
   "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>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>tags</th>\n",
       "      <th>max_rating</th>\n",
       "      <th>Rank_advanced</th>\n",
       "      <th>Rank_beginner</th>\n",
       "      <th>Rank_expert</th>\n",
       "      <th>Rank_intermediate</th>\n",
       "      <th>Level_A</th>\n",
       "      <th>...</th>\n",
       "      <th>number theory</th>\n",
       "      <th>probabilities</th>\n",
       "      <th>schedules</th>\n",
       "      <th>shortest paths</th>\n",
       "      <th>sortings</th>\n",
       "      <th>string suffix structures</th>\n",
       "      <th>strings</th>\n",
       "      <th>ternary search</th>\n",
       "      <th>trees</th>\n",
       "      <th>two pointers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>strings</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_5781</td>\n",
       "      <td>1</td>\n",
       "      <td>implementation,math</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_137</td>\n",
       "      <td>2</td>\n",
       "      <td>implementation</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1550</td>\n",
       "      <td>2</td>\n",
       "      <td>brute force,dfs and similar</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1322</td>\n",
       "      <td>3</td>\n",
       "      <td>dfs and similar,shortest paths</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 59 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id problem_id  attempts_range                            tags  \\\n",
       "0  user_232  prob_6507               1                         strings   \n",
       "1  user_232  prob_5781               1             implementation,math   \n",
       "2  user_232   prob_137               2                  implementation   \n",
       "3  user_232  prob_1550               2     brute force,dfs and similar   \n",
       "4  user_232  prob_1322               3  dfs and similar,shortest paths   \n",
       "\n",
       "   max_rating  Rank_advanced  Rank_beginner  Rank_expert  Rank_intermediate  \\\n",
       "0     307.913              0              1            0                  0   \n",
       "1     307.913              0              1            0                  0   \n",
       "2     307.913              0              1            0                  0   \n",
       "3     307.913              0              1            0                  0   \n",
       "4     307.913              0              1            0                  0   \n",
       "\n",
       "   Level_A      ...       number theory  probabilities  schedules  \\\n",
       "0        0      ...                   0              0          0   \n",
       "1        1      ...                   0              0          0   \n",
       "2        0      ...                   0              0          0   \n",
       "3        1      ...                   0              0          0   \n",
       "4        0      ...                   0              0          0   \n",
       "\n",
       "   shortest paths  sortings  string suffix structures  strings  \\\n",
       "0               0         0                         0        1   \n",
       "1               0         0                         0        0   \n",
       "2               0         0                         0        0   \n",
       "3               0         0                         0        0   \n",
       "4               1         0                         0        0   \n",
       "\n",
       "   ternary search  trees  two pointers  \n",
       "0               0      0             0  \n",
       "1               0      0             0  \n",
       "2               0      0             0  \n",
       "3               0      0             0  \n",
       "4               0      0             0  \n",
       "\n",
       "[5 rows x 59 columns]"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "train=train.drop(['tags'],axis=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "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>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "      <th>max_rating</th>\n",
       "      <th>Rank_advanced</th>\n",
       "      <th>Rank_beginner</th>\n",
       "      <th>Rank_expert</th>\n",
       "      <th>Rank_intermediate</th>\n",
       "      <th>Level_A</th>\n",
       "      <th>Level_B</th>\n",
       "      <th>...</th>\n",
       "      <th>number theory</th>\n",
       "      <th>probabilities</th>\n",
       "      <th>schedules</th>\n",
       "      <th>shortest paths</th>\n",
       "      <th>sortings</th>\n",
       "      <th>string suffix structures</th>\n",
       "      <th>strings</th>\n",
       "      <th>ternary search</th>\n",
       "      <th>trees</th>\n",
       "      <th>two pointers</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_6507</td>\n",
       "      <td>1</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_5781</td>\n",
       "      <td>1</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_137</td>\n",
       "      <td>2</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1550</td>\n",
       "      <td>2</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_232</td>\n",
       "      <td>prob_1322</td>\n",
       "      <td>3</td>\n",
       "      <td>307.913</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>...</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 58 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "    user_id problem_id  attempts_range  max_rating  Rank_advanced  \\\n",
       "0  user_232  prob_6507               1     307.913              0   \n",
       "1  user_232  prob_5781               1     307.913              0   \n",
       "2  user_232   prob_137               2     307.913              0   \n",
       "3  user_232  prob_1550               2     307.913              0   \n",
       "4  user_232  prob_1322               3     307.913              0   \n",
       "\n",
       "   Rank_beginner  Rank_expert  Rank_intermediate  Level_A  Level_B  \\\n",
       "0              1            0                  0        0        1   \n",
       "1              1            0                  0        1        0   \n",
       "2              1            0                  0        0        1   \n",
       "3              1            0                  0        1        0   \n",
       "4              1            0                  0        0        1   \n",
       "\n",
       "       ...       number theory  probabilities  schedules  shortest paths  \\\n",
       "0      ...                   0              0          0               0   \n",
       "1      ...                   0              0          0               0   \n",
       "2      ...                   0              0          0               0   \n",
       "3      ...                   0              0          0               0   \n",
       "4      ...                   0              0          0               1   \n",
       "\n",
       "   sortings  string suffix structures  strings  ternary search  trees  \\\n",
       "0         0                         0        1               0      0   \n",
       "1         0                         0        0               0      0   \n",
       "2         0                         0        0               0      0   \n",
       "3         0                         0        0               0      0   \n",
       "4         0                         0        0               0      0   \n",
       "\n",
       "   two pointers  \n",
       "0             0  \n",
       "1             0  \n",
       "2             0  \n",
       "3             0  \n",
       "4             0  \n",
       "\n",
       "[5 rows x 58 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "train['user_id']=train['user_id'].str.replace('user_','')\n",
    "train['problem_id']=train['problem_id'].str.replace('prob_','')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/weight_boosting.py:29: DeprecationWarning: numpy.core.umath_tests is an internal NumPy module and should not be imported. It will be removed in a future NumPy release.\n",
      "  from numpy.core.umath_tests import inner1d\n"
     ]
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.ensemble.gradient_boosting import GradientBoostingClassifier\n",
    "from sklearn.metrics import f1_score\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "from sklearn.linear_model import LogisticRegression, LogisticRegressionCV\n",
    "from sklearn.model_selection import train_test_split"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "#function for calculating the F1 score\n",
    "def compute_score(y_true, y_pred, average='weighted'):\n",
    "    xval = f1_score(y_true, y_pred, average=average)\n",
    "    return np.mean(xval)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "X=train[['user_id','problem_id','max_rating','Rank_advanced','Rank_beginner','Rank_expert','Rank_intermediate','Level_A','Level_B','Level_C',\n",
    " 'Level_D','Level_E','Level_F','Level_G','Level_H','Level_I','Level_J','Level_K','Level_L','Level_M','Level_N','*special','2-sat',\n",
    " 'binary search','bitmasks','brute force','chinese remainder theorem','combinatorics','constructive algorithms','data structures','dfs and similar',\n",
    " 'divide and conquer','dp','dsu','expression parsing','fft','flows','games','geometry','graph matchings','graphs','greedy','hashing',\n",
    " 'implementation','math','matrices','meet-in-the-middle','number theory','probabilities','schedules','shortest paths','sortings','string suffix structures',\n",
    " 'strings','ternary search','trees','two pointers']]\n",
    "y=train[['attempts_range']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "#splitting the training and testing data\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.33, random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:2: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  \n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "0.5035708710583827"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf = RandomForestClassifier()\n",
    "clf = clf.fit(X_train, y_train)\n",
    "clf.score(X_test,y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "#calculating the importance of each feature on the attempts_range\n",
    "features=pd.DataFrame(clf.feature_importances_,\n",
    "                                   index = X_train.columns,\n",
    "                                    columns=['importance']).sort_values('importance',ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a368cda20>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1800x1800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "features.plot(kind='barh', figsize=(25, 25))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(104047, 3)\n"
     ]
    }
   ],
   "source": [
    "#making changes to the dataset on the bases of feature's importance\n",
    "model = SelectFromModel(clf, prefit=True)\n",
    "train_reduced = model.transform(X_train)\n",
    "print (train_reduced.shape)\n",
    "test_reduced=model.transform(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:578: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score Validation of : <class 'sklearn.linear_model.logistic.LogisticRegression'>\n",
      "F1 Score score = 0.3735229684708074\n",
      "****\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/metrics/classification.py:1135: UndefinedMetricWarning: F-score is ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score Validation of : <class 'sklearn.linear_model.logistic.LogisticRegressionCV'>\n",
      "F1 Score score = 0.3735229684708074\n",
      "****\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:8: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  \n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score Validation of : <class 'sklearn.ensemble.forest.RandomForestClassifier'>\n",
      "F1 Score score = 0.43016500493050625\n",
      "****\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py:578: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples, ), for example using ravel().\n",
      "  y = column_or_1d(y, warn=True)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "F1 Score Validation of : <class 'sklearn.ensemble.gradient_boosting.GradientBoostingClassifier'>\n",
      "F1 Score score = 0.3891285821706229\n",
      "****\n"
     ]
    }
   ],
   "source": [
    "#Calculating the F1 score\n",
    "logreg = LogisticRegression()\n",
    "logreg_cv = LogisticRegressionCV()\n",
    "rf = RandomForestClassifier()\n",
    "gboost = GradientBoostingClassifier()\n",
    "\n",
    "models = [logreg, logreg_cv, rf, gboost]\n",
    "for model in models:\n",
    "    model=model.fit(train_reduced, y_train)\n",
    "    prediction=pd.DataFrame(model.predict(test_reduced))\n",
    "    print ('F1 Score Validation of : {0}'.format(model.__class__))\n",
    "    score = compute_score(y_true=y_test, y_pred=prediction, average='weighted')\n",
    "    print ('F1 Score score = {0}'.format(score))\n",
    "    print ('****')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:2: DataConversionWarning: A column-vector y was passed when a 1d array was expected. Please change the shape of y to (n_samples,), for example using ravel().\n",
      "  \n"
     ]
    }
   ],
   "source": [
    "#Since We are getting the highest score in RFC so will go forward with that\n",
    "rf=rf.fit(train_reduced, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 66555 entries, 0 to 66554\n",
      "Data columns (total 3 columns):\n",
      "ID            66555 non-null object\n",
      "user_id       66555 non-null object\n",
      "problem_id    66555 non-null object\n",
      "dtypes: object(3)\n",
      "memory usage: 1.5+ MB\n"
     ]
    }
   ],
   "source": [
    "test.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "#checking for duplicate entries in the test sheet\n",
    "test= test.drop_duplicates(subset=['ID'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 66555 entries, 0 to 66554\n",
      "Data columns (total 3 columns):\n",
      "ID            66555 non-null object\n",
      "user_id       66555 non-null object\n",
      "problem_id    66555 non-null object\n",
      "dtypes: object(3)\n",
      "memory usage: 2.0+ MB\n"
     ]
    }
   ],
   "source": [
    "test.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_prediction_columns=test[['user_id','problem_id']]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_prediction_columns=pd.merge(test_prediction_columns,user_data[['user_id','max_rating']],on='user_id')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_prediction_columns['user_id']=test_prediction_columns['user_id'].str.replace('user_','')\n",
    "test_prediction_columns['problem_id']=test_prediction_columns['problem_id'].str.replace('prob_','')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [],
   "source": [
    "#making the prediction\n",
    "output=rf.predict(test_prediction_columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "test['attempts_range']=output"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "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>ID</th>\n",
       "      <th>user_id</th>\n",
       "      <th>problem_id</th>\n",
       "      <th>attempts_range</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>user_856_prob_5822</td>\n",
       "      <td>user_856</td>\n",
       "      <td>prob_5822</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>user_2642_prob_2334</td>\n",
       "      <td>user_2642</td>\n",
       "      <td>prob_2334</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>user_2557_prob_2920</td>\n",
       "      <td>user_2557</td>\n",
       "      <td>prob_2920</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>user_1572_prob_4598</td>\n",
       "      <td>user_1572</td>\n",
       "      <td>prob_4598</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>user_295_prob_6139</td>\n",
       "      <td>user_295</td>\n",
       "      <td>prob_6139</td>\n",
       "      <td>2</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    ID    user_id problem_id  attempts_range\n",
       "0   user_856_prob_5822   user_856  prob_5822               1\n",
       "1  user_2642_prob_2334  user_2642  prob_2334               3\n",
       "2  user_2557_prob_2920  user_2557  prob_2920               2\n",
       "3  user_1572_prob_4598  user_1572  prob_4598               1\n",
       "4   user_295_prob_6139   user_295  prob_6139               2"
      ]
     },
     "execution_count": 85,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "#saving the results to a csv file\n",
    "test[['ID','user_id','problem_id','attempts_range']].to_csv('prediction.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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}