NLP Based Price Recommendation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/sparshtekriwal/opt/anaconda3/lib/python3.7/site-packages/statsmodels/tools/_testing.py:19: FutureWarning: pandas.util.testing is deprecated. Use the functions in the public API at pandas.testing instead.\n",
      "  import pandas.util.testing as tm\n"
     ]
    }
   ],
   "source": [
    "## Importing required libraries\n",
    "import pandas as pd\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import warnings\n",
    "import pickle\n",
    "import gc\n",
    "warnings.filterwarnings(\"ignore\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking for variables in memory "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                           _i2:  601.0 B\n",
      "                           _oh:  248.0 B\n",
      "                           Out:  248.0 B\n",
      "                            _i:  247.0 B\n",
      "                           _i1:  247.0 B\n",
      "                    sizeof_fmt:  144.0 B\n",
      "                       __doc__:  113.0 B\n",
      "                           _ih:  104.0 B\n",
      "                            In:  104.0 B\n",
      "                   __builtin__:   88.0 B\n",
      "                  __builtins__:   88.0 B\n",
      "                            pd:   88.0 B\n",
      "                            np:   88.0 B\n",
      "                           plt:   88.0 B\n",
      "                           sns:   88.0 B\n",
      "                      warnings:   88.0 B\n",
      "                        pickle:   88.0 B\n",
      "                            gc:   88.0 B\n",
      "                           sys:   88.0 B\n",
      "                           _dh:   80.0 B\n"
     ]
    }
   ],
   "source": [
    "gc.collect()\n",
    "import sys\n",
    "def sizeof_fmt(num, suffix='B'):\n",
    "    ''' by Fred Cirera,  https://stackoverflow.com/a/1094933/1870254, modified'''\n",
    "    for unit in ['','Ki','Mi','Gi','Ti','Pi','Ei','Zi']:\n",
    "        if abs(num) < 1024.0:\n",
    "            return \"%3.1f %s%s\" % (num, unit, suffix)\n",
    "        num /= 1024.0\n",
    "    return \"%.1f %s%s\" % (num, 'Yi', suffix)\n",
    "\n",
    "\n",
    "for name, size in sorted(((name, sys.getsizeof(value)) for name, value in locals().items()),\n",
    "                         key= lambda x: -x[1])[:20]:\n",
    "    print(\"{:>30}: {:>8}\".format(name, sizeof_fmt(size)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Reading File and Scaling Numerical Predictors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Before Scaling Condition ID\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>name</th>\n",
       "      <th>item_condition_id</th>\n",
       "      <th>category_name</th>\n",
       "      <th>brand_name</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>item_description</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MLB Cincinnati Reds T Shirt Size XL</td>\n",
       "      <td>3</td>\n",
       "      <td>Men/Tops/T-shirts</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>No description yet</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Razer BlackWidow Chroma Keyboard</td>\n",
       "      <td>3</td>\n",
       "      <td>Electronics/Computers &amp; Tablets/Components &amp; P...</td>\n",
       "      <td>Razer</td>\n",
       "      <td>52.0</td>\n",
       "      <td>0</td>\n",
       "      <td>This keyboard is in great condition and works ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AVA-VIV Blouse</td>\n",
       "      <td>1</td>\n",
       "      <td>Women/Tops &amp; Blouses/Blouse</td>\n",
       "      <td>Target</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>Adorable top with a hint of lace and a key hol...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Leather Horse Statues</td>\n",
       "      <td>1</td>\n",
       "      <td>Home/Home Décor/Home Décor Accents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>New with tags. Leather horses. Retail for [rm]...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24K GOLD plated rose</td>\n",
       "      <td>1</td>\n",
       "      <td>Women/Jewelry/Necklaces</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>Complete with certificate of authenticity</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name  item_condition_id  \\\n",
       "train_id                                                           \n",
       "0         MLB Cincinnati Reds T Shirt Size XL                  3   \n",
       "1            Razer BlackWidow Chroma Keyboard                  3   \n",
       "2                              AVA-VIV Blouse                  1   \n",
       "3                       Leather Horse Statues                  1   \n",
       "4                        24K GOLD plated rose                  1   \n",
       "\n",
       "                                              category_name brand_name  price  \\\n",
       "train_id                                                                        \n",
       "0                                         Men/Tops/T-shirts        NaN   10.0   \n",
       "1         Electronics/Computers & Tablets/Components & P...      Razer   52.0   \n",
       "2                               Women/Tops & Blouses/Blouse     Target   10.0   \n",
       "3                        Home/Home Décor/Home Décor Accents        NaN   35.0   \n",
       "4                                   Women/Jewelry/Necklaces        NaN   44.0   \n",
       "\n",
       "          shipping                                   item_description  \n",
       "train_id                                                               \n",
       "0                1                                 No description yet  \n",
       "1                0  This keyboard is in great condition and works ...  \n",
       "2                1  Adorable top with a hint of lace and a key hol...  \n",
       "3                1  New with tags. Leather horses. Retail for [rm]...  \n",
       "4                0          Complete with certificate of authenticity  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "After Scaling Condition ID\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>name</th>\n",
       "      <th>item_condition_id</th>\n",
       "      <th>category_name</th>\n",
       "      <th>brand_name</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>item_description</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MLB Cincinnati Reds T Shirt Size XL</td>\n",
       "      <td>0.5</td>\n",
       "      <td>Men/Tops/T-shirts</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>No description yet</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Razer BlackWidow Chroma Keyboard</td>\n",
       "      <td>0.5</td>\n",
       "      <td>Electronics/Computers &amp; Tablets/Components &amp; P...</td>\n",
       "      <td>Razer</td>\n",
       "      <td>52.0</td>\n",
       "      <td>0</td>\n",
       "      <td>This keyboard is in great condition and works ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AVA-VIV Blouse</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Women/Tops &amp; Blouses/Blouse</td>\n",
       "      <td>Target</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>Adorable top with a hint of lace and a key hol...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Leather Horse Statues</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Home/Home Décor/Home Décor Accents</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>New with tags. Leather horses. Retail for [rm]...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24K GOLD plated rose</td>\n",
       "      <td>0.0</td>\n",
       "      <td>Women/Jewelry/Necklaces</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>Complete with certificate of authenticity</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name  item_condition_id  \\\n",
       "train_id                                                           \n",
       "0         MLB Cincinnati Reds T Shirt Size XL                0.5   \n",
       "1            Razer BlackWidow Chroma Keyboard                0.5   \n",
       "2                              AVA-VIV Blouse                0.0   \n",
       "3                       Leather Horse Statues                0.0   \n",
       "4                        24K GOLD plated rose                0.0   \n",
       "\n",
       "                                              category_name brand_name  price  \\\n",
       "train_id                                                                        \n",
       "0                                         Men/Tops/T-shirts        NaN   10.0   \n",
       "1         Electronics/Computers & Tablets/Components & P...      Razer   52.0   \n",
       "2                               Women/Tops & Blouses/Blouse     Target   10.0   \n",
       "3                        Home/Home Décor/Home Décor Accents        NaN   35.0   \n",
       "4                                   Women/Jewelry/Necklaces        NaN   44.0   \n",
       "\n",
       "          shipping                                   item_description  \n",
       "train_id                                                               \n",
       "0                1                                 No description yet  \n",
       "1                0  This keyboard is in great condition and works ...  \n",
       "2                1  Adorable top with a hint of lace and a key hol...  \n",
       "3                1  New with tags. Leather horses. Retail for [rm]...  \n",
       "4                0          Complete with certificate of authenticity  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Reading File as df\n",
    "df = pd.read_csv(\"train.tsv\",sep=\"\\t\", nrows=10000)\n",
    "# df_copy = df\n",
    "df.index=df[\"train_id\"]\n",
    "del df[\"train_id\"]\n",
    "print(\"Before Scaling Condition ID\")\n",
    "display(df.head(5))\n",
    "\n",
    "\n",
    "# Scaling Item condition column\n",
    "from sklearn.preprocessing import MinMaxScaler as mms\n",
    "# Scaling feature for linear models\n",
    "df[\"item_condition_id\"]=mms().fit_transform(df[\"item_condition_id\"].to_numpy().reshape(-1,1))\n",
    "print(\"After Scaling Condition ID\")\n",
    "display(df.head(5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Exploration and Preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Avg number of words in the feature--name: 4.4077\n",
      "Avg number of words in the feature--item_description: 25.6353\n"
     ]
    }
   ],
   "source": [
    "for col in [\"name\", \"item_description\"]:\n",
    "    print(\"Avg number of words in the feature--%s: %s\"%(col, df[col].apply(lambda x: len(str(x).split())).mean()))"
   ]
  },
  {
   "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>Missing Values</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>name</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>item_condition_id</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>category_name</th>\n",
       "      <td>47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>brand_name</th>\n",
       "      <td>4261</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>price</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>shipping</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>item_description</th>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   Missing Values\n",
       "name                            0\n",
       "item_condition_id               0\n",
       "category_name                  47\n",
       "brand_name                   4261\n",
       "price                           0\n",
       "shipping                        0\n",
       "item_description                0"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(df.isna().sum(), columns=[\"Missing Values\"])"
   ]
  },
  {
   "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>name</th>\n",
       "      <th>item_condition_id</th>\n",
       "      <th>brand_name</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>item_description</th>\n",
       "      <th>general_cat</th>\n",
       "      <th>subcat_1</th>\n",
       "      <th>subcat_2</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>MLB Cincinnati Reds T Shirt Size XL</td>\n",
       "      <td>0.50</td>\n",
       "      <td>NaN</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>No description yet</td>\n",
       "      <td>Men</td>\n",
       "      <td>Tops</td>\n",
       "      <td>T-shirts</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>Razer BlackWidow Chroma Keyboard</td>\n",
       "      <td>0.50</td>\n",
       "      <td>Razer</td>\n",
       "      <td>52.0</td>\n",
       "      <td>0</td>\n",
       "      <td>This keyboard is in great condition and works ...</td>\n",
       "      <td>Electronics</td>\n",
       "      <td>Computers &amp; Tablets</td>\n",
       "      <td>Components &amp; Parts</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>AVA-VIV Blouse</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Target</td>\n",
       "      <td>10.0</td>\n",
       "      <td>1</td>\n",
       "      <td>Adorable top with a hint of lace and a key hol...</td>\n",
       "      <td>Women</td>\n",
       "      <td>Tops &amp; Blouses</td>\n",
       "      <td>Blouse</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Leather Horse Statues</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>35.0</td>\n",
       "      <td>1</td>\n",
       "      <td>New with tags. Leather horses. Retail for [rm]...</td>\n",
       "      <td>Home</td>\n",
       "      <td>Home Décor</td>\n",
       "      <td>Home Décor Accents</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>24K GOLD plated rose</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>44.0</td>\n",
       "      <td>0</td>\n",
       "      <td>Complete with certificate of authenticity</td>\n",
       "      <td>Women</td>\n",
       "      <td>Jewelry</td>\n",
       "      <td>Necklaces</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9995</th>\n",
       "      <td>Halloween nail stickers</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>3.0</td>\n",
       "      <td>1</td>\n",
       "      <td>New. Both sets included *please only purchase ...</td>\n",
       "      <td>Beauty</td>\n",
       "      <td>Makeup</td>\n",
       "      <td>Nails</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9996</th>\n",
       "      <td>NWT Wildfox Fringe Top XS Coachella</td>\n",
       "      <td>0.00</td>\n",
       "      <td>Wildfox Couture</td>\n",
       "      <td>17.0</td>\n",
       "      <td>0</td>\n",
       "      <td>NWT Wildfox Coachella Fringe Top size XS. Mint...</td>\n",
       "      <td>Women</td>\n",
       "      <td>Tops &amp; Blouses</td>\n",
       "      <td>T-Shirts</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9997</th>\n",
       "      <td>Like New Baby Carrier</td>\n",
       "      <td>0.25</td>\n",
       "      <td>Chicco</td>\n",
       "      <td>15.0</td>\n",
       "      <td>1</td>\n",
       "      <td>Excellent condition. No tears no stains used m...</td>\n",
       "      <td>Kids</td>\n",
       "      <td>Gear</td>\n",
       "      <td>Backpacks &amp; Carriers</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9998</th>\n",
       "      <td>Minecraft Creeper Scarf</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>12.0</td>\n",
       "      <td>1</td>\n",
       "      <td>A scarf is a trusty companion, covering your n...</td>\n",
       "      <td>Kids</td>\n",
       "      <td>Boys (4+)</td>\n",
       "      <td>Accessories</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9999</th>\n",
       "      <td>Michael Kors Swim Top</td>\n",
       "      <td>0.25</td>\n",
       "      <td>Michael Kors</td>\n",
       "      <td>11.0</td>\n",
       "      <td>0</td>\n",
       "      <td>MICHAEL Michael Kors Halter Logo Tankini Top t...</td>\n",
       "      <td>Women</td>\n",
       "      <td>Swimwear</td>\n",
       "      <td>One-Piece</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10000 rows × 9 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                         name  item_condition_id  \\\n",
       "train_id                                                           \n",
       "0         MLB Cincinnati Reds T Shirt Size XL               0.50   \n",
       "1            Razer BlackWidow Chroma Keyboard               0.50   \n",
       "2                              AVA-VIV Blouse               0.00   \n",
       "3                       Leather Horse Statues               0.00   \n",
       "4                        24K GOLD plated rose               0.00   \n",
       "...                                       ...                ...   \n",
       "9995                  Halloween nail stickers               0.00   \n",
       "9996      NWT Wildfox Fringe Top XS Coachella               0.00   \n",
       "9997                    Like New Baby Carrier               0.25   \n",
       "9998                  Minecraft Creeper Scarf               0.00   \n",
       "9999                    Michael Kors Swim Top               0.25   \n",
       "\n",
       "               brand_name  price  shipping  \\\n",
       "train_id                                     \n",
       "0                     NaN   10.0         1   \n",
       "1                   Razer   52.0         0   \n",
       "2                  Target   10.0         1   \n",
       "3                     NaN   35.0         1   \n",
       "4                     NaN   44.0         0   \n",
       "...                   ...    ...       ...   \n",
       "9995                  NaN    3.0         1   \n",
       "9996      Wildfox Couture   17.0         0   \n",
       "9997               Chicco   15.0         1   \n",
       "9998                  NaN   12.0         1   \n",
       "9999         Michael Kors   11.0         0   \n",
       "\n",
       "                                           item_description  general_cat  \\\n",
       "train_id                                                                   \n",
       "0                                        No description yet          Men   \n",
       "1         This keyboard is in great condition and works ...  Electronics   \n",
       "2         Adorable top with a hint of lace and a key hol...        Women   \n",
       "3         New with tags. Leather horses. Retail for [rm]...         Home   \n",
       "4                 Complete with certificate of authenticity        Women   \n",
       "...                                                     ...          ...   \n",
       "9995      New. Both sets included *please only purchase ...       Beauty   \n",
       "9996      NWT Wildfox Coachella Fringe Top size XS. Mint...        Women   \n",
       "9997      Excellent condition. No tears no stains used m...         Kids   \n",
       "9998      A scarf is a trusty companion, covering your n...         Kids   \n",
       "9999      MICHAEL Michael Kors Halter Logo Tankini Top t...        Women   \n",
       "\n",
       "                     subcat_1              subcat_2  \n",
       "train_id                                             \n",
       "0                        Tops              T-shirts  \n",
       "1         Computers & Tablets    Components & Parts  \n",
       "2              Tops & Blouses                Blouse  \n",
       "3                  Home Décor    Home Décor Accents  \n",
       "4                     Jewelry             Necklaces  \n",
       "...                       ...                   ...  \n",
       "9995                   Makeup                 Nails  \n",
       "9996           Tops & Blouses              T-Shirts  \n",
       "9997                     Gear  Backpacks & Carriers  \n",
       "9998                Boys (4+)           Accessories  \n",
       "9999                 Swimwear             One-Piece  \n",
       "\n",
       "[10000 rows x 9 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# splitting Category Women/Jewelry/Necklaces into three distinct columns general_cat, subcat_1, subcat_2 and storing it back to data frame\n",
    "    \n",
    "def split_categories(df):\n",
    "    new_df = df.copy()\n",
    "    new_df[\"category_name\"].replace(to_replace=\"Unknown\",value=\"Unknown/Unknown/Unknown\", inplace=True)\n",
    "    new_df[\"category_name\"].fillna(\"Unknown/Unknown/Unknown\",inplace=True)\n",
    "    #Splitting the category_name into 3 categories:\n",
    "    new_df['general_cat'], new_df['subcat_1'], new_df['subcat_2'] = zip(*new_df['category_name'].apply(lambda x: x.split(\"/\")))\n",
    "    new_df.drop([\"category_name\"], axis=1, inplace=True)\n",
    "    return new_df\n",
    "\n",
    "df = split_categories(df)\n",
    "display(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- There is alot of variance in the Price column which is our target variable\n",
    "- The histogram is skewed to the Right\n",
    "- After taking log of the price column, we get a distribution that looks normally distributed\n",
    "- Hence we will be using log transformation our output column and adding one to deal with 0 values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(1, 2, 1)\n",
    "(df['price']).plot.hist(bins=50, figsize=(20,10), edgecolor='white',range=[0,250])\n",
    "plt.xlabel('Price', fontsize=17)\n",
    "plt.ylabel('frequency', fontsize=17)\n",
    "plt.tick_params(labelsize=15)\n",
    "plt.text(125, 1500, r'$\\mu=\\$ 26.73, \\sigma=\\$ 38.58 $', fontsize=17)\n",
    "plt.grid()\n",
    "plt.title('Price Distribution (Left skewed)', fontsize=17)\n",
    "\n",
    "plt.subplot(1, 2, 2)\n",
    "np.log1p(df['price']).plot.hist(bins=50, figsize=(20,10), edgecolor='white')\n",
    "plt.xlabel('log(Price+1)', fontsize=17)\n",
    "plt.ylabel('frequency', fontsize=17)\n",
    "plt.text(5, 800, r'$\\mu= 2.97 , \\sigma= 0.749 $', fontsize=17)\n",
    "plt.tick_params(labelsize=15)\n",
    "plt.title('Log(Price) Distribution (Normal) ', fontsize=17)\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean log price: 2.96904885842168\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean log price: \"+ str(np.log1p(df['price']).mean()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Mean log price: 0.7419182565636051\n"
     ]
    }
   ],
   "source": [
    "print(\"Mean log price: \"+ str(np.log1p(df['price']).std()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "scrolled": true
   },
   "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>item_condition_id</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "      <td>10000.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.226250</td>\n",
       "      <td>26.429900</td>\n",
       "      <td>0.453900</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.227027</td>\n",
       "      <td>39.385515</td>\n",
       "      <td>0.497895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>10.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.250000</td>\n",
       "      <td>16.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>0.500000</td>\n",
       "      <td>29.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1506.000000</td>\n",
       "      <td>1.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       item_condition_id         price      shipping\n",
       "count       10000.000000  10000.000000  10000.000000\n",
       "mean            0.226250     26.429900      0.453900\n",
       "std             0.227027     39.385515      0.497895\n",
       "min             0.000000      0.000000      0.000000\n",
       "25%             0.000000     10.000000      0.000000\n",
       "50%             0.250000     16.000000      0.000000\n",
       "75%             0.500000     29.000000      1.000000\n",
       "max             1.000000   1506.000000      1.000000"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Int64Index: 10000 entries, 0 to 9999\n",
      "Data columns (total 9 columns):\n",
      " #   Column             Non-Null Count  Dtype  \n",
      "---  ------             --------------  -----  \n",
      " 0   name               10000 non-null  object \n",
      " 1   item_condition_id  10000 non-null  float64\n",
      " 2   brand_name         5739 non-null   object \n",
      " 3   price              10000 non-null  float64\n",
      " 4   shipping           10000 non-null  int64  \n",
      " 5   item_description   10000 non-null  object \n",
      " 6   general_cat        10000 non-null  object \n",
      " 7   subcat_1           10000 non-null  object \n",
      " 8   subcat_2           10000 non-null  object \n",
      "dtypes: float64(2), int64(1), object(6)\n",
      "memory usage: 781.2+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### No signs of multicolianrity with the numerical columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "\n",
    "plt.figure(figsize=(9,6))\n",
    "sns.heatmap(df.corr(), annot=True, cmap=\"viridis\")\n",
    "plt.xticks([0,1,2,3])\n",
    "plt.yticks([0,1,2,3])\n",
    "plt.tight_layout()\n",
    "plt.title(\"Corelation Heatmap\", fontsize=17)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting the Top N brands irrespective of category"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1008x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Nike                 350\n",
      "PINK                 348\n",
      "Victoria's Secret    322\n",
      "LuLaRoe              201\n",
      "Apple                111\n",
      "FOREVER 21           108\n",
      "Nintendo              99\n",
      "Michael Kors          87\n",
      "Sephora               86\n",
      "Bath & Body Works     85\n",
      "Lululemon             85\n",
      "Rae Dunn              80\n",
      "Adidas                77\n",
      "Disney                72\n",
      "American Eagle        68\n",
      "Coach                 66\n",
      "Under Armour          60\n",
      "Old Navy              59\n",
      "Too Faced             52\n",
      "Sony                  48\n",
      "Name: brand_name, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "def plot_topn_brands(topn):\n",
    "    \n",
    "    frequently_listed_brands=list(df[\"brand_name\"].value_counts().iloc[:topn].index)\n",
    "    value_counts = list(df[\"brand_name\"].value_counts().iloc[:topn].values)\n",
    "    value_counts = [\"n: \" + str(i) for i in value_counts]\n",
    "    fig = plt.figure(1, figsize=(14, 8))\n",
    "    data = df[df[\"brand_name\"].isin(frequently_listed_brands)]\n",
    "    sns.boxplot(x=\"price\", y=\"brand_name\", data=data, fliersize=2, palette=\"coolwarm\")\n",
    "    plt.title(\"Price distribution of top %s brands by frequency\"%(topn), fontsize=25)\n",
    "    plt.xlim(0,200)\n",
    "    plt.yticks(fontsize=15)\n",
    "    plt.xticks(fontsize=15)\n",
    "    plt.xlabel(\"Price\", fontsize=20)\n",
    "    plt.ylabel(\"Brand Name\", fontsize=20)\n",
    "    plt.show()\n",
    "    print(df[\"brand_name\"].value_counts()[:topn])\n",
    "    \n",
    "plot_topn_brands(20)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Motherhood Maternity    4\n",
       "Keurig                  4\n",
       "Oakley                  4\n",
       "Similac                 4\n",
       "Star Wars               4\n",
       "                       ..\n",
       "Longchamp               1\n",
       "Ben Nye                 1\n",
       "Pacific Sunwear         1\n",
       "Cartier                 1\n",
       "TopShop                 1\n",
       "Name: brand_name, Length: 561, dtype: int64"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[\"brand_name\"].value_counts()[200:]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "There are 11 unique geenral categories.\n",
      "There are 106 unique first sub-categories.\n",
      "There are 478 unique second sub-categories.\n",
      "There are 762 unique brand names in the training dataset.\n"
     ]
    }
   ],
   "source": [
    "print(\"There are %d unique geenral categories.\" % df['general_cat'].nunique())\n",
    "print(\"There are %d unique first sub-categories.\" % df['subcat_1'].nunique())\n",
    "print(\"There are %d unique second sub-categories.\" % df['subcat_2'].nunique())\n",
    "print(\"There are %d unique brand names in the training dataset.\" % df['brand_name'].apply(lambda x:str(x).lower()).nunique())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### One Hot Encoding Categorical Features - Categories and Brands"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "- As some brands have very few products listed, we will only consider the top n brands as features for our model and put the others into another group called \"non-frequent brands\"\n",
    "- \"combine_category_and_brand_vectors\" function returns a sparse martix containing the vectorized brands and Categories column list will help keep track of the original features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Takes df and topn as parameter and returns a dataframe\n",
    "def vectorize_brand_name(df, topn=-1):\n",
    "    frequently_listed_brands=list(df[\"brand_name\"].value_counts().iloc[:topn].index)\n",
    "    brand_df =  pd.get_dummies(df[\"brand_name\"].apply(lambda x: \"non-frequent brands\" if x not in frequently_listed_brands else x), prefix=\"Brand\")\n",
    "    return brand_df.drop([\"Brand_non-frequent brands\"], axis=1)\n",
    "# vectorize_brand_name(df).head(10)\n",
    "\n",
    "# Taking all 11 categories and returning a dataframe Gen_Cat_name. Unknown category is omitted\n",
    "def vectorize_general_cat(df, topn=11):\n",
    "    frequently_listed_categories=list(df[\"general_cat\"].value_counts().iloc[:topn].index)\n",
    "    category_df =  pd.get_dummies(df[\"general_cat\"].apply(lambda x: \"non-frequent categories\" if x not in frequently_listed_categories else x), prefix=\"Gen_Cat\")\n",
    "    try:\n",
    "        return category_df.drop([\"Gen_Cat_Unknown\",\"non-frequent categories\"], axis=1)\n",
    "    except:\n",
    "        return category_df.drop([\"Gen_Cat_Unknown\"], axis=1)\n",
    "# vectorize_general_cat(df).head(10)    \n",
    "\n",
    "# Taking top n sub categories 1 out of 106 and returning a dataframe Sub_Cat_1_name\n",
    "def vectorize_sub_cat_1(df, topn=-1):\n",
    "    frequently_listed_categories=list(df[\"subcat_1\"].value_counts().iloc[:topn].index)\n",
    "    category_df =  pd.get_dummies(df[\"subcat_1\"].apply(lambda x: \"non-frequent categories\" if x not in frequently_listed_categories else x), prefix=\"Sub_Cat_1\")\n",
    "    try:\n",
    "        return category_df.drop([\"Sub_Cat_1_Unknown\", \"Sub_Cat_1_non-frequent categories\"], axis=1)\n",
    "    except:\n",
    "        return category_df.drop([\"Sub_Cat_1_Unknown\"], axis=1)\n",
    "# vectorize_sub_cat_1(df).head(10)    \n",
    "\n",
    "# Taking top n sub categories 2 out of 762 and returning a dataframe Sub_Cat_2_name\n",
    "def vectorize_sub_cat_2(df, topn=-1):\n",
    "    frequently_listed_categories=list(df[\"subcat_2\"].value_counts().iloc[:topn].index)\n",
    "    category_df =  pd.get_dummies(df[\"subcat_2\"].apply(lambda x: \"non-frequent categories\" if x not in frequently_listed_categories else x), prefix=\"Sub_Cat_2\")\n",
    "    try:\n",
    "        return category_df.drop([\"Sub_Cat_2_Unknown\", \"Sub_Cat_2_non-frequent categories\"], axis=1)\n",
    "    except:\n",
    "        return category_df.drop([\"Sub_Cat_2_Unknown\"], axis=1)\n",
    "# vectorize_sub_cat_2(df).head(10)  \n",
    "\n",
    "\n",
    "def combine_category_and_brand_vectors(dataFrame):\n",
    "    from scipy.sparse import csr_matrix, hstack\n",
    "    brands = vectorize_brand_name(dataFrame)\n",
    "    gen_cat = vectorize_general_cat(dataFrame)\n",
    "    sub_cat_1 = vectorize_sub_cat_1(dataFrame)\n",
    "    sub_cat_2 = vectorize_sub_cat_2(dataFrame)\n",
    "    sparse_array_cat_brands =  hstack(( csr_matrix(brands), csr_matrix(gen_cat), csr_matrix(sub_cat_1), csr_matrix(sub_cat_2) )).tocsr()\n",
    "    column_list = list(brands.columns)+list(gen_cat.columns)+list(sub_cat_1.columns)+list(sub_cat_2.columns)\n",
    "    return sparse_array_cat_brands, column_list\n",
    "\n",
    "\n",
    "# array_cat_brand, column_cat_brand = combine_category_and_brand_vectors(df[df[\"lemmatized_text\"]!=\"\"])\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Checking for variables in memory "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                            df:  6.9 MiB\n",
      "                             _: 61.0 KiB\n",
      "                           _14: 61.0 KiB\n",
      "                            _i:  2.9 KiB\n",
      "                          _i16:  2.9 KiB\n",
      "                           mms:  1.2 KiB\n",
      "                           ___: 1012.0 B\n",
      "                            _5: 1012.0 B\n",
      "                          _i13:  823.0 B\n",
      "                           _i7:  743.0 B\n",
      "                           _i6:  714.0 B\n",
      "                            __:  707.0 B\n",
      "                           _10:  707.0 B\n",
      "                          _i17:  603.0 B\n",
      "                           _i2:  601.0 B\n",
      "                           _i3:  510.0 B\n",
      "                           _ii:  410.0 B\n",
      "                          _i15:  410.0 B\n",
      "                           _ih:  272.0 B\n",
      "                            In:  272.0 B\n"
     ]
    }
   ],
   "source": [
    "# gc.collect()\n",
    "import sys\n",
    "def sizeof_fmt(num, suffix='B'):\n",
    "    ''' by Fred Cirera,  https://stackoverflow.com/a/1094933/1870254, modified'''\n",
    "    for unit in ['','Ki','Mi','Gi','Ti','Pi','Ei','Zi']:\n",
    "        if abs(num) < 1024.0:\n",
    "            return \"%3.1f %s%s\" % (num, unit, suffix)\n",
    "        num /= 1024.0\n",
    "    return \"%.1f %s%s\" % (num, 'Yi', suffix)\n",
    "\n",
    "\n",
    "for name, size in sorted(((name, sys.getsizeof(value)) for name, value in locals().items()),\n",
    "                         key= lambda x: -x[1])[:20]:\n",
    "    print(\"{:>30}: {:>8}\".format(name, sizeof_fmt(size)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Processing Text Columns "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Combining Name and Description Columns \n",
    "- Converting to lower case to avoid duplicates\n",
    "- Removing dummy text \"No description yet\" from description feature as we are making the \"assumption\" that a missing descrition doesn't impact the price "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"unprocessed_text\"] =  df[\"name\"] +\" \"+ df[\"item_description\"].replace(to_replace=\"No description yet\",value=\"\")\n",
    "df[\"unprocessed_text\"] = df[\"unprocessed_text\"].apply(lambda x: str(x).lower())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Counts of words as tuples before preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('chroma', 1),\n",
       " ('synapse', 1),\n",
       " ('ava-viv', 1),\n",
       " ('statues', 1),\n",
       " ('horses.', 1),\n",
       " ('ruie', 1),\n",
       " ('blazer,amy', 1),\n",
       " ('byers', 1),\n",
       " ('tides', 1),\n",
       " ('santorini', 1),\n",
       " ('shortened', 1),\n",
       " ('tumbling', 1),\n",
       " ('sophie', 1),\n",
       " ('bra/boy', 1),\n",
       " ('store!)', 1),\n",
       " ('checker', 1),\n",
       " ('backwards', 1),\n",
       " ('clowns', 1),\n",
       " ('scary.', 1),\n",
       " ('0.25', 1),\n",
       " ('pi', 1),\n",
       " ('dogs2016', 1),\n",
       " ('5.5-7', 1),\n",
       " ('winter)', 1),\n",
       " ('.05oz/1.5g', 1),\n",
       " ('.034oz/1ml', 1),\n",
       " ('eyeshadows;', 1),\n",
       " (\"k'tan\", 1),\n",
       " ('k’tan', 1),\n",
       " ('hi-tech', 1),\n",
       " ('sweat,', 1),\n",
       " ('sun’s', 1),\n",
       " ('rays,', 1),\n",
       " ('development.', 1),\n",
       " ('double-loop', 1),\n",
       " ('\"merry', 1),\n",
       " ('macaroons\"', 1),\n",
       " ('the.', 1),\n",
       " ('sized\"better', 1),\n",
       " ('sex\"', 1),\n",
       " ('impeccable', 1),\n",
       " ('cream/', 1),\n",
       " ('fancy,', 1),\n",
       " ('polyester;', 1),\n",
       " ('dried.', 1),\n",
       " ('anthropology,', 1),\n",
       " ('goods,', 1),\n",
       " ('-maximum', 1),\n",
       " ('(unlocked)', 1),\n",
       " ('welcomed.', 1),\n",
       " ('you...', 1),\n",
       " ('lanascloset', 1),\n",
       " ('delete/update', 1),\n",
       " ('“sold”', 1),\n",
       " ('carmine', 1),\n",
       " ('29w.', 1),\n",
       " ('33l.', 1),\n",
       " ('leathers', 1),\n",
       " ('tore', 1),\n",
       " ('qty3', 1),\n",
       " ('4.8g', 1),\n",
       " ('/0.17', 1),\n",
       " ('band!', 1),\n",
       " ('caudalíe', 1),\n",
       " ('caudalie', 1),\n",
       " ('pugsley', 1),\n",
       " ('skirted', 1),\n",
       " ('pinksuper', 1),\n",
       " ('martini', 1),\n",
       " ('marabou', 1),\n",
       " ('#kitschoure', 1),\n",
       " ('#beasweetlollipopinaworldofsoursuckers', 1),\n",
       " ('diego', 1),\n",
       " ('cali-fornia', 1),\n",
       " ('a!', 1),\n",
       " ('foxy', 1),\n",
       " ('loxy', 1),\n",
       " ('dollface!', 1),\n",
       " ('girlfriend.', 1),\n",
       " ('rockabilly', 1),\n",
       " ('edf', 1),\n",
       " ('raver', 1),\n",
       " ('1in', 1),\n",
       " ('togo', 1),\n",
       " ('twilly.', 1),\n",
       " ('downloaded', 1),\n",
       " ('gameplay', 1),\n",
       " ('expos', 1),\n",
       " ('buh', 1),\n",
       " ('bra\"', 1),\n",
       " ('feb..', 1),\n",
       " ('apricot', 1),\n",
       " ('foundation!!', 1),\n",
       " ('quality!!!', 1),\n",
       " ('bowl✨', 1),\n",
       " ('sticky,', 1),\n",
       " ('scemted', 1),\n",
       " ('*let', 1),\n",
       " ('*usually', 1),\n",
       " ('purchase...', 1),\n",
       " ('concern.••', 1),\n",
       " ('mascara...', 1),\n",
       " ('300%...', 1),\n",
       " ('turquaz,', 1),\n",
       " ('wheels-super', 1),\n",
       " ('tuners', 1),\n",
       " ('\"carl\\'s', 1),\n",
       " ('collectibles\".', 1),\n",
       " ('year:', 1),\n",
       " ('mattel,', 1),\n",
       " ('inc.', 1),\n",
       " ('\"favorite', 1),\n",
       " ('cars\"', 1),\n",
       " ('rids', 1),\n",
       " ('wood.', 1),\n",
       " ('broach.all', 1),\n",
       " ('drip', 1),\n",
       " ('yokai', 1),\n",
       " ('sweatproof', 1),\n",
       " ('headsets', 1),\n",
       " ('ring!', 1),\n",
       " ('solitaires.', 1),\n",
       " ('long),', 1),\n",
       " ('-rarely', 1),\n",
       " ('-selling', 1),\n",
       " ('-35%', 1),\n",
       " ('37%', 1),\n",
       " ('18%', 1),\n",
       " ('angora', 1),\n",
       " ('❕please', 1),\n",
       " ('ordering❕', 1),\n",
       " ('❕❕❕if', 1),\n",
       " ('cancel❕❕❕', 1),\n",
       " ('⏺️provides', 1),\n",
       " ('drops!', 1),\n",
       " ('⏺️price', 1),\n",
       " ('⏺️brand', 1),\n",
       " ('⏺️will', 1),\n",
       " ('⏺️check', 1),\n",
       " ('⏺️allow', 1),\n",
       " ('update!!', 1),\n",
       " ('ultraboost', 1),\n",
       " ('partners', 1),\n",
       " ('shipfromchina', 1),\n",
       " ('\"fine', 1),\n",
       " ('fashion:', 1),\n",
       " ('\"shipping', 1),\n",
       " ('china,competitive', 1),\n",
       " ('days~', 1),\n",
       " ('aknuckles1', 1),\n",
       " ('flameless', 1),\n",
       " ('melt/tart/essential', 1),\n",
       " ('watt', 1),\n",
       " ('~3.5', 1),\n",
       " ('house,', 1),\n",
       " ('ranch,', 1),\n",
       " ('cabin,', 1),\n",
       " ('ie.', 1),\n",
       " ('pkgs', 1),\n",
       " ('chips!', 1),\n",
       " ('terminated', 1),\n",
       " ('fluorite:', 1),\n",
       " ('\"protection', 1),\n",
       " ('stone\"', 1),\n",
       " ('spiritualizing', 1),\n",
       " ('wishing', 1),\n",
       " ('wicca', 1),\n",
       " ('artifact', 1),\n",
       " ('witchcraftartifact', 1),\n",
       " ('once;', 1),\n",
       " ('*hold*free', 1),\n",
       " ('today-it', 1),\n",
       " ('kern.', 1),\n",
       " ('products...and', 1),\n",
       " ('meandering', 1),\n",
       " ('products...search', 1),\n",
       " ('plussizemakeup', 1),\n",
       " ('listings...you', 1),\n",
       " ('#plussizemakeup', 1),\n",
       " ('medium;', 1),\n",
       " ('directions,', 1),\n",
       " ('espresso', 1),\n",
       " ('pink/reddish', 1),\n",
       " ('sheet!', 1),\n",
       " ('bed!', 1),\n",
       " ('jillian', 1),\n",
       " ('michaels', 1),\n",
       " ('plan,', 1),\n",
       " ('rotational', 1),\n",
       " ('calendar.', 1),\n",
       " ('ignite,', 1),\n",
       " ('launch,', 1),\n",
       " ('amplify,', 1),\n",
       " ('escalate,', 1),\n",
       " ('conquer,', 1),\n",
       " ('triumph,', 1),\n",
       " ('zenith,', 1),\n",
       " ('apex.', 1),\n",
       " ('opus', 1),\n",
       " ('ignore:', 1),\n",
       " ('p90x', 1),\n",
       " ('doug', 1),\n",
       " ('folding', 1),\n",
       " ('tentacruel', 1),\n",
       " ('6y.', 1),\n",
       " ('jewel', 1),\n",
       " ('goosebumps', 1),\n",
       " ('horrorland', 1),\n",
       " ('(8),', 1),\n",
       " ('draggle', 1),\n",
       " ('vhtf!', 1),\n",
       " ('{size', 1),\n",
       " ('8}', 1),\n",
       " ('dry,', 1),\n",
       " ('khakis', 1),\n",
       " ('luminess', 1),\n",
       " ('groupon', 1),\n",
       " ('error!', 1),\n",
       " ('how-to', 1),\n",
       " ('\"medium', 1),\n",
       " ('shade\"', 1),\n",
       " ('match\",', 1),\n",
       " ('case!!', 1),\n",
       " ('trip!', 1),\n",
       " ('compliments!!', 1),\n",
       " ('ears!', 1),\n",
       " ('neck!', 1),\n",
       " ('bezel', 1),\n",
       " ('stickiness', 1),\n",
       " ('whatsoever.', 1),\n",
       " ('bled', 1),\n",
       " ('distance.', 1),\n",
       " ('intact,', 1),\n",
       " ('tacos', 1),\n",
       " ('tequila?', 1),\n",
       " ('treats.', 1),\n",
       " ('heaven,', 1),\n",
       " ('fall/winter,', 1),\n",
       " ('shorter.!!!', 1),\n",
       " ('bathingsuit', 1),\n",
       " ('kesha', 1),\n",
       " ('kesha:', 1),\n",
       " ('slider', 1),\n",
       " ('wedges!', 1),\n",
       " ('pain??', 1),\n",
       " ('electrical', 1),\n",
       " ('nerves', 1),\n",
       " ('modify', 1),\n",
       " ('perception.', 1),\n",
       " ('works/what', 1),\n",
       " ('pain...', 1),\n",
       " ('simulate', 1),\n",
       " ('nerve(s).', 1),\n",
       " ('sensation.', 1),\n",
       " ('postings.', 1),\n",
       " ('busty', 1),\n",
       " ('seperately,', 1),\n",
       " ('lifts', 1),\n",
       " ('sewed.', 1),\n",
       " ('sizes=', 1),\n",
       " ('spartina', 1),\n",
       " ('fobs', 1),\n",
       " ('straps/', 1),\n",
       " ('clasps.', 1),\n",
       " ('franco', 1),\n",
       " ('sarto', 1),\n",
       " ('tapestry!', 1),\n",
       " ('88\"', 1),\n",
       " ('55\"', 1),\n",
       " ('recordings!', 1),\n",
       " ('plosive', 1),\n",
       " ('\"\"p\"\"', 1),\n",
       " ('\"\"b\"\"', 1),\n",
       " ('sudden', 1),\n",
       " ('bassy', 1),\n",
       " ('\"\"pop\"\"', 1),\n",
       " ('recordings,', 1),\n",
       " ('ruining', 1),\n",
       " ('clamping', 1),\n",
       " ('bolt', 1),\n",
       " ('clam', 1),\n",
       " ('saliva/spit', 1),\n",
       " ('15cm', 1),\n",
       " ('35cm', 1),\n",
       " ('297g', 1),\n",
       " ('tigger.', 1),\n",
       " ('mentha', 1),\n",
       " ('snap/', 1),\n",
       " ('divider,', 1),\n",
       " ('15\".', 1),\n",
       " ('superstar', 1),\n",
       " ('adjstraps', 1),\n",
       " ('designer~', 1),\n",
       " ('details~', 1),\n",
       " ('size~', 1),\n",
       " ('condition~', 1),\n",
       " ('(outgrew', 1),\n",
       " ('chance)', 1),\n",
       " ('**3', 1),\n",
       " ('promo,', 1),\n",
       " ('sellout:', 1),\n",
       " ('nikki3promo', 1),\n",
       " ('bar**', 1),\n",
       " ('thanks**', 1),\n",
       " ('#easter', 1),\n",
       " ('#anytimedress', 1),\n",
       " ('#nikki3promo', 1),\n",
       " ('amiibo.', 1),\n",
       " ('manolo', 1),\n",
       " ('blanik', 1),\n",
       " (\"3'or\", 1),\n",
       " ('dotted', 1),\n",
       " ('bouse', 1),\n",
       " ('bumi,', 1),\n",
       " ('bounces', 1),\n",
       " ('floats', 1),\n",
       " ('safe!', 1),\n",
       " ('breds', 1),\n",
       " ('sleep!', 1),\n",
       " ('(#2)', 1),\n",
       " ('schunk', 1),\n",
       " ('set;', 1),\n",
       " ('bracket.', 1),\n",
       " ('5c.', 1),\n",
       " ('outter', 1),\n",
       " ('ever-faithful', 1),\n",
       " ('york®', 1),\n",
       " ('cobble', 1),\n",
       " ('shoulder/elbow', 1),\n",
       " ('foil-stamped', 1),\n",
       " ('upright', 1),\n",
       " ('bookstripe', 1),\n",
       " ('back-wall', 1),\n",
       " ('\"*depth:', 1),\n",
       " ('1⁄4', 1),\n",
       " ('\"*height:', 1),\n",
       " ('mustard,', 1),\n",
       " ('iredesent', 1),\n",
       " (')(◕‿◕)', 1),\n",
       " ('fishbowl,', 1),\n",
       " ('color)=^_^=', 1),\n",
       " ('(╹◡╹)', 1),\n",
       " ('slime)', 1),\n",
       " ('!!**', 1),\n",
       " ('⊙︿⊙', 1),\n",
       " ('aferg16.', 1),\n",
       " ('thuy', 1),\n",
       " ('1999', 1),\n",
       " ('grader,', 1),\n",
       " ('deliver.', 1),\n",
       " ('xlg', 1),\n",
       " ('esteem,', 1),\n",
       " ('asks', 1),\n",
       " ('games;', 1),\n",
       " ('baublebar', 1),\n",
       " ('bm', 1),\n",
       " ('bargain!', 1),\n",
       " ('tranquilo', 1),\n",
       " ('awesome,', 1),\n",
       " ('fussy', 1),\n",
       " ('babies,', 1),\n",
       " ('colic.', 1),\n",
       " ('etc...no', 1),\n",
       " ('pest', 1),\n",
       " ('eveshadow', 1),\n",
       " ('bending.', 1),\n",
       " ('duces', 1),\n",
       " ('90%.', 1),\n",
       " ('cables?', 1),\n",
       " ('batman,', 1),\n",
       " ('america,', 1),\n",
       " ('scrump,', 1),\n",
       " ('dale,', 1),\n",
       " ('squirrel,', 1),\n",
       " ('paw,', 1),\n",
       " ('stitch,', 1),\n",
       " ('sully,', 1),\n",
       " ('mike,', 1),\n",
       " ('inc.,', 1),\n",
       " ('kitty,', 1),\n",
       " ('owl,', 1),\n",
       " ('chopper,', 1),\n",
       " (\"chi's\", 1),\n",
       " ('anime,', 1),\n",
       " ('amw', 1),\n",
       " ('bench', 1),\n",
       " ('2k14', 1),\n",
       " ('1/2zip', 1),\n",
       " ('(phone', 1),\n",
       " ('rubber.', 1),\n",
       " ('absorbent,', 1),\n",
       " ('shatterproof,', 1),\n",
       " ('grease,', 1),\n",
       " ('abrasions,', 1),\n",
       " ('voiding', 1),\n",
       " ('warranty.', 1),\n",
       " ('serve:', 1),\n",
       " ('desperate', 1),\n",
       " ('funds.', 1),\n",
       " ('auth.', 1),\n",
       " ('s-8\"', 1),\n",
       " ('bracelet.box', 1),\n",
       " ('400%*', 1),\n",
       " ('lashes+', 1),\n",
       " ('****firm', 1),\n",
       " ('price****', 1),\n",
       " ('marucci', 1),\n",
       " ('33/30', 1),\n",
       " ('logan', 1),\n",
       " ('xxsp', 1),\n",
       " ('ship✨', 1),\n",
       " ('welcomed,', 1),\n",
       " ('colourpop!', 1),\n",
       " ('(sunday)', 1),\n",
       " ('may.', 1),\n",
       " (',retail', 1),\n",
       " ('(headband', 1),\n",
       " ('baby).', 1),\n",
       " ('rosette', 1),\n",
       " ('sashes.', 1),\n",
       " ('shoot!', 1),\n",
       " ('78\".', 1),\n",
       " ('brightness,', 1),\n",
       " ('revel', 1),\n",
       " ('colada', 1),\n",
       " ('teapot.', 1),\n",
       " ('inventory!', 1),\n",
       " ('(men', 1),\n",
       " ('style).', 1),\n",
       " ('moving!', 1),\n",
       " ('kettle', 1),\n",
       " ('order..', 1),\n",
       " ('bin#32', 1),\n",
       " ('remove.', 1),\n",
       " ('oso', 1),\n",
       " ('goofy,', 1),\n",
       " ('signatures', 1),\n",
       " ('mesh/velvet', 1),\n",
       " ('personally,', 1),\n",
       " ('slot,', 1),\n",
       " ('3.75in', 1),\n",
       " ('9.5cm', 1),\n",
       " ('declined.', 1),\n",
       " ('dinner,', 1),\n",
       " ('dustbag,', 1),\n",
       " ('16.25\"', 1),\n",
       " ('used)', 1),\n",
       " ('dents.', 1),\n",
       " ('fl.oz.liq.', 1),\n",
       " ('4-6?', 1),\n",
       " ('125lbs', 1),\n",
       " ('2805.', 1),\n",
       " ('bodice.', 1),\n",
       " ('floor-length', 1),\n",
       " ('floor.', 1),\n",
       " ('glitter/shimmer', 1),\n",
       " ('located.', 1),\n",
       " ('berkley', 1),\n",
       " ('lures', 1),\n",
       " ('chameleon', 1),\n",
       " ('sexier', 1),\n",
       " ('lrresustible', 1),\n",
       " ('wag', 1),\n",
       " ('pause', 1),\n",
       " ('sll', 1),\n",
       " ('wright:', 1),\n",
       " ('8g', 1),\n",
       " ('2in', 1),\n",
       " ('walking!', 1),\n",
       " ('-shyterria', 1),\n",
       " ('nx', 1),\n",
       " ('(xbox', 1),\n",
       " ('iv', 1),\n",
       " ('redemption', 1),\n",
       " ('goty', 1),\n",
       " ('•bnwt', 1),\n",
       " ('•new,', 1),\n",
       " ('•shipped', 1),\n",
       " ('dog-friendly', 1),\n",
       " ('hand/shoulder', 1),\n",
       " ('roomy.15x10.5x4.', 1),\n",
       " ('e10+', 1),\n",
       " ('warner', 1),\n",
       " ('zagg', 1),\n",
       " ('#145', 1),\n",
       " ('tracing', 1),\n",
       " ('ae.', 1),\n",
       " ('choices', 1),\n",
       " ('costs!', 1),\n",
       " ('teals', 1),\n",
       " ('stretchier', 1),\n",
       " ('forver', 1),\n",
       " ('regenessence', 1),\n",
       " ('sized,', 1),\n",
       " ('multi-firming', 1),\n",
       " ('*see', 1),\n",
       " ('10-13', 1),\n",
       " ('bubblegum', 1),\n",
       " ('classy...', 1),\n",
       " ('flavor.', 1),\n",
       " ('24g', 1),\n",
       " ('overstock!', 1),\n",
       " ('straight-leg', 1),\n",
       " ('w/turquoise', 1),\n",
       " ('ricky', 1),\n",
       " ('garner', 1),\n",
       " ('(black/purple)', 1),\n",
       " ('coupon!', 1),\n",
       " ('bag(free', 1),\n",
       " ('perfecr', 1),\n",
       " ('occassion.it', 1),\n",
       " (\"one.don't\", 1),\n",
       " ('interested.inappropriate', 1),\n",
       " ('deleted/blocked!', 1),\n",
       " ('remotes', 1),\n",
       " ('rekindle', 1),\n",
       " ('swag773', 1),\n",
       " ('longsleeve', 1),\n",
       " ('sixteen', 1),\n",
       " ('41\"', 1),\n",
       " ('giving.', 1),\n",
       " ('trainer/xsmall/ships', 1),\n",
       " ('morning!!!newest', 1),\n",
       " ('(2017)', 1),\n",
       " ('trainer....message', 1),\n",
       " ('post...', 1),\n",
       " ('....3rd', 1),\n",
       " ('trainer.high', 1),\n",
       " ('4/', 1),\n",
       " ('changes.if', 1),\n",
       " ('post....i', 1),\n",
       " ('2x,3x,4x', 1),\n",
       " ('costume!', 1),\n",
       " ('*stretchy', 1),\n",
       " ('*fuzzy', 1),\n",
       " ('*light', 1),\n",
       " ('items*', 1),\n",
       " ('fallin\"', 1),\n",
       " ('stork', 1),\n",
       " ('storks', 1),\n",
       " ('reveals!', 1),\n",
       " ('quickly...', 1),\n",
       " ('outlaw', 1),\n",
       " ('maison.', 1),\n",
       " ('#l', 1),\n",
       " ('garland.', 1),\n",
       " ('(0.24', 1),\n",
       " ('hydrating,', 1),\n",
       " ('super-charged', 1),\n",
       " ('particles', 1),\n",
       " ('blur,', 1),\n",
       " ('reflect,', 1),\n",
       " ('dullness', 1),\n",
       " ('freya', 1),\n",
       " ('sparrows', 1),\n",
       " ('❗any', 1),\n",
       " ('ignored❗', 1),\n",
       " ('crocs.', 1),\n",
       " ('5.8oz', 1),\n",
       " ('versailles', 1),\n",
       " ('confetti', 1),\n",
       " ('thai', 1),\n",
       " ('beautuful', 1),\n",
       " ('1.2oz', 1),\n",
       " ('fragrances,', 1),\n",
       " ('beige/taupe', 1),\n",
       " ('top?', 1),\n",
       " ('worries', 1),\n",
       " ('much/little', 1),\n",
       " ('boobage.', 1),\n",
       " ('wilfred', 1),\n",
       " ('chested', 1),\n",
       " ('**new', 1),\n",
       " ('**very', 1),\n",
       " ('**silver', 1),\n",
       " ('sand.', 1),\n",
       " ('*hermione', 1),\n",
       " ('granger', 1),\n",
       " ('*wizards', 1),\n",
       " ('*fantasy', 1),\n",
       " ('present!!!', 1),\n",
       " ('discounts:', 1),\n",
       " ('added!', 1),\n",
       " ('rudolph', 1),\n",
       " ('nosed', 1),\n",
       " ('reindeer', 1),\n",
       " ('tags:,', 1),\n",
       " ('lula,', 1),\n",
       " ('nwt/', 1),\n",
       " ('makeup:)', 1),\n",
       " ('separate!:)', 1),\n",
       " ('-instant', 1),\n",
       " ('840', 1),\n",
       " ('❤☺nyx', 1),\n",
       " ('✨✨✨choose', 1),\n",
       " ('[rm]✨✨✨', 1),\n",
       " ('✨✨please', 1),\n",
       " ('closet...thanks!!✨✨', 1),\n",
       " ('⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆⬆', 1),\n",
       " ('❤**all', 1),\n",
       " ('photographed', 1),\n",
       " ('shipment**❤', 1),\n",
       " ('✨✨discount', 1),\n",
       " ('bundles✨✨', 1),\n",
       " ('❤☺(1)', 1),\n",
       " ('⭐brand', 1),\n",
       " ('sealed⭐', 1),\n",
       " ('✨highly', 1),\n",
       " ('pigmented✨', 1),\n",
       " ('✨high', 1),\n",
       " ('quality✨', 1),\n",
       " ('✨long', 1),\n",
       " ('lasting✨', 1),\n",
       " ('☺**price', 1),\n",
       " ('firm**☺', 1),\n",
       " ('secret/vs', 1),\n",
       " ('liners/pencils', 1),\n",
       " ('eye/dip', 1),\n",
       " (\"j's\", 1),\n",
       " ('day!!❤☺❤', 1),\n",
       " ('rt-500', 1),\n",
       " ('compact,', 1),\n",
       " ('stowe.', 1),\n",
       " ('!ship', 1),\n",
       " ('ok??', 1),\n",
       " ('colorescience', 1),\n",
       " ('htf!!!', 1),\n",
       " ('butt,', 1),\n",
       " ('legs,', 1),\n",
       " ('vs/vs', 1),\n",
       " ('cheaper.', 1),\n",
       " ('nephew', 1),\n",
       " ('20inches', 1),\n",
       " ('table.', 1),\n",
       " ('spooky,', 1),\n",
       " ('witches,', 1),\n",
       " ('spiders,', 1),\n",
       " ('cobwebs,', 1),\n",
       " ('pumpkins,', 1),\n",
       " ('weeks!', 1),\n",
       " ('#eyelashes', 1),\n",
       " ('#longlashes', 1),\n",
       " ('#growth', 1),\n",
       " ('#rejuvenator', 1),\n",
       " ('#thick', 1),\n",
       " ('#serum', 1),\n",
       " ('#eye', 1),\n",
       " ('#eyemakeup', 1),\n",
       " ('#eyelashextensions', 1),\n",
       " ('#safe', 1),\n",
       " ('#sephora', 1),\n",
       " ('savor', 1),\n",
       " ('sig', 1),\n",
       " ('carton', 1),\n",
       " ('kawaii,', 1),\n",
       " ('supply,', 1),\n",
       " ('nivea', 1),\n",
       " ('lhall', 1),\n",
       " ('-l', 1),\n",
       " ('top-', 1),\n",
       " ('brazil-authentic.', 1),\n",
       " ('■▪', 1),\n",
       " ('▪■', 1),\n",
       " ('▪no', 1),\n",
       " ('▪size', 1),\n",
       " ('▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪▪', 1),\n",
       " ('urbanoutfitters', 1),\n",
       " ('modcloth', 1),\n",
       " ('lipsense.', 1),\n",
       " ('ragland', 1),\n",
       " ('white/burgundy', 1),\n",
       " ('grey-royal', 1),\n",
       " ('white-soverign', 1),\n",
       " ('seas,', 1),\n",
       " ('blake', 1),\n",
       " ('27\".', 1),\n",
       " ('lovestruck', 1),\n",
       " ('.12', 1),\n",
       " ('lightening', 1),\n",
       " ('.17', 1),\n",
       " ('foils:', 1),\n",
       " ('t9', 1),\n",
       " ('wolverine,', 1),\n",
       " ('skeletor,', 1),\n",
       " ('cape),', 1),\n",
       " ('boy!!', 1),\n",
       " ('cleats/spikes/shoes', 1),\n",
       " (\"men's/boys\", 1),\n",
       " ('spikes', 1),\n",
       " ('rubber/plastic', 1),\n",
       " ('cleat', 1),\n",
       " ('reflected.', 1),\n",
       " ('length\"', 1),\n",
       " ('23.5', 1),\n",
       " ('preowned.', 1),\n",
       " ('32x30', 1),\n",
       " ('medium!!', 1),\n",
       " ('overstocked', 1),\n",
       " ('locations)', 1),\n",
       " ('subtract', 1),\n",
       " ('ruler:', 1),\n",
       " ('20.5\"', 1),\n",
       " ('donotbuy!', 1),\n",
       " ('bazaar', 1),\n",
       " ('-camouflage', 1),\n",
       " ('freddy', 1),\n",
       " ('comment**', 1),\n",
       " ('hh', 1),\n",
       " ('9022', 1),\n",
       " (\"5's\", 1),\n",
       " (\"6's+ipads\", 1),\n",
       " ('myself,', 1),\n",
       " ('interested!!', 1),\n",
       " ('length:3', 1),\n",
       " ('type:8', 1),\n",
       " ('weight:0.4', 1),\n",
       " ('apple:iphone', 1),\n",
       " ('guaranteed–this', 1),\n",
       " ('faux-fur', 1),\n",
       " ('60\"w', 1),\n",
       " ('72\"l', 1),\n",
       " ('mvmt', 1),\n",
       " ('canyon', 1),\n",
       " ('blues,', 1),\n",
       " ('red/coral', 1),\n",
       " ('###no', 1),\n",
       " ('holds###.', 1),\n",
       " ('things!', 1),\n",
       " ('**shipping', 1),\n",
       " ('(dropping', 1),\n",
       " ('office)', 1),\n",
       " ('monday-saturday.', 1),\n",
       " ('shopping!!', 1),\n",
       " ('storing-', 1),\n",
       " ('pulls-', 1),\n",
       " ('snowmobile', 1),\n",
       " ('headbands,', 1),\n",
       " ('magaschoni', 1),\n",
       " ('cashmere,', 1),\n",
       " ('khakis,', 1),\n",
       " ('panties,', 1),\n",
       " ('citronella/gray', 1),\n",
       " ('camo.', 1),\n",
       " ('this\"', 1),\n",
       " ('loungefly', 1),\n",
       " ('canvas/faux', 1),\n",
       " ('loungefly.', 1),\n",
       " ('kitty.', 1),\n",
       " ('imaginext,', 1),\n",
       " ('wars,', 1),\n",
       " ('few.', 1),\n",
       " ('ios.', 1),\n",
       " ('watches!', 1),\n",
       " ('iwatch.', 1),\n",
       " ('warren', 1),\n",
       " ('7m', 1),\n",
       " ('friendship', 1),\n",
       " ('dec.', 1),\n",
       " ('10kt', 1),\n",
       " ('incase', 1),\n",
       " ('line!', 1),\n",
       " ('lip-color', 1),\n",
       " ('spontaneous.', 1),\n",
       " ('seal!', 1),\n",
       " ('0.493', 1),\n",
       " ('goat-milk', 1),\n",
       " ('camera!', 1),\n",
       " ('by:maya.', 1),\n",
       " ('tritan', 1),\n",
       " ('each(not', 1),\n",
       " ('lowering!', 1),\n",
       " ('splitting!', 1),\n",
       " ('substitutions!', 1),\n",
       " ('***asking=blocked!', 1),\n",
       " ('5+', 1),\n",
       " ('exceptions!', 1),\n",
       " ('wanelo,', 1),\n",
       " ('****ask', 1),\n",
       " ('purchasing!****', 1),\n",
       " ('days,usps', 1),\n",
       " ('lot❗️', 1),\n",
       " ('shipped,due', 1),\n",
       " ('fault)❗️', 1),\n",
       " ('➡️so', 1),\n",
       " ('me.⬅️', 1),\n",
       " ('-bundle', 1),\n",
       " ('-returning', 1),\n",
       " ('❌holds|❌try', 1),\n",
       " ('!|❌trades', 1),\n",
       " ('♡♡♡', 1),\n",
       " ('come,first', 1),\n",
       " ('intentional', 1),\n",
       " (\"'welcome\", 1),\n",
       " (\"nevada'\", 1),\n",
       " ('-kids', 1),\n",
       " ('\"angel\"', 1),\n",
       " ('loungewear.', 1),\n",
       " ('\\'garbage\"', 1),\n",
       " ('garbage', 1),\n",
       " ('greek', 1),\n",
       " ('crème', 1),\n",
       " ('intensity.', 1),\n",
       " ('créme', 1),\n",
       " ('a.e.', 1),\n",
       " ('markings.', 1),\n",
       " ('jaclyn', 1),\n",
       " ('seahawks', 1),\n",
       " ('‼️‼️shipped', 1),\n",
       " ('✨✨✨please', 1),\n",
       " ('you✨✨', 1),\n",
       " ('smearing', 1),\n",
       " ('❤️permanent', 1),\n",
       " ('❤️impossible', 1),\n",
       " ('sets)', 1),\n",
       " ('sets-', 1),\n",
       " ('konichiwa', 1),\n",
       " ('pallettes', 1),\n",
       " ('brights', 1),\n",
       " ('nothing.', 1),\n",
       " ('jleman', 1),\n",
       " ('waist,slightly', 1),\n",
       " ('(42', 1),\n",
       " ('1/2,', 1),\n",
       " ('jersey,', 1),\n",
       " ('age,', 1),\n",
       " ('lateralus', 1),\n",
       " ('nave', 1),\n",
       " ('uwd', 1),\n",
       " ('curves!', 1),\n",
       " ('wardrobe!!', 1),\n",
       " ('goodcondition', 1),\n",
       " ('mexico,', 1),\n",
       " ('slidewatch!', 1),\n",
       " ('decorating.', 1),\n",
       " ('happiness.', 1),\n",
       " ('1daysale', 1),\n",
       " ('cuties', 1),\n",
       " ('snags,', 1),\n",
       " ('7w', 1),\n",
       " ('engravemybones', 1),\n",
       " ('09', 1),\n",
       " (\"skellington's\", 1),\n",
       " ('stains/no', 1),\n",
       " ('w/minor', 1),\n",
       " ('sweaters/hoodies.', 1),\n",
       " ('walks', 1),\n",
       " ('jr.interactive', 1),\n",
       " ('see,', 1),\n",
       " ('hear,', 1),\n",
       " ('dance!', 1),\n",
       " ('forever!', 1),\n",
       " ('describes', 1),\n",
       " ('program', 1),\n",
       " ('tricks…etc.', 1),\n",
       " ('internet', 1),\n",
       " ('sensors', 1),\n",
       " ('sensor,', 1),\n",
       " ('sensor.', 1),\n",
       " ('‘emotions’', 1),\n",
       " ('“thunder', 1),\n",
       " ('electronics', 1),\n",
       " ('…made', 1),\n",
       " ('”', 1),\n",
       " ('(fixable', 1),\n",
       " ('polish),', 1),\n",
       " ('insole,', 1),\n",
       " ('7....', 1),\n",
       " ('condition....', 1),\n",
       " ('bamboo’s', 1),\n",
       " ('plant-based:', 1),\n",
       " ('wrapper,', 1),\n",
       " ('62%', 1),\n",
       " ('(bio-based', 1),\n",
       " ('biodegradable).', 1),\n",
       " ('(organic', 1),\n",
       " ('pesticides', 1),\n",
       " ('fertilizers).', 1),\n",
       " ('plant-based', 1),\n",
       " ('commercially', 1),\n",
       " ('compostable.', 1),\n",
       " ('encasing', 1),\n",
       " ('usda', 1),\n",
       " ('biobased.', 1),\n",
       " ('certified.', 1),\n",
       " ('vegan.', 1),\n",
       " ('bpa-free.', 1),\n",
       " ('non-toxic.', 1),\n",
       " ('masquerade', 1),\n",
       " ('15/16.', 1),\n",
       " ('•ucla', 1),\n",
       " ('sweatshirt•', 1),\n",
       " ('▪️youth', 1),\n",
       " ('▪️unisex', 1),\n",
       " ('▪️super', 1),\n",
       " ('▪️signature', 1),\n",
       " ('▪️50%', 1),\n",
       " ('▪️chest', 1),\n",
       " ('▪️length', 1),\n",
       " ('#130', 1),\n",
       " ('3!!!', 1),\n",
       " ('boy/girl', 1),\n",
       " ('manufacturer*', 1),\n",
       " ('*quality', 1),\n",
       " ('assured', 1),\n",
       " ('direct*', 1),\n",
       " ('mattifies', 1),\n",
       " ('corrects', 1),\n",
       " (\"✔you'll\", 1),\n",
       " ('✅handling', 1),\n",
       " ('\"ships', 1),\n",
       " ('within\"', 1),\n",
       " ('means)', 1),\n",
       " ('✔take', 1),\n",
       " ('✔each', 1),\n",
       " ('product)', 1),\n",
       " ('❌our', 1),\n",
       " ('highlighters,', 1),\n",
       " ('jason.', 1),\n",
       " ('color/vampy', 1),\n",
       " ('red/burgundy', 1),\n",
       " ('box:)', 1),\n",
       " ('ops', 1),\n",
       " ('pollack', 1),\n",
       " ('\"liquid', 1),\n",
       " ('silver\".', 1),\n",
       " ('mornings!', 1),\n",
       " ('generous', 1),\n",
       " ('stretchy..one', 1),\n",
       " ('sepia', 1),\n",
       " ('swatch,', 1),\n",
       " ('belkin', 1),\n",
       " ('chargers!', 1),\n",
       " ('salvage', 1),\n",
       " ('auction', 1),\n",
       " ('recall', 1),\n",
       " ('from,', 1),\n",
       " ('medium/large.', 1),\n",
       " ('5\\'4\"', 1),\n",
       " ('ware.', 1),\n",
       " ('w/tracking', 1),\n",
       " ('♡bnib', 1),\n",
       " ('♡unbranded,', 1),\n",
       " ('♡prices', 1),\n",
       " ('•firm•', 1),\n",
       " ('lashes!', 1),\n",
       " ('♡i', 1),\n",
       " ('mst.', 1),\n",
       " ('♡excludes', 1),\n",
       " ('holidays/weekends', 1),\n",
       " ('views:', 1),\n",
       " ('velour,', 1),\n",
       " ('mink,', 1),\n",
       " ('wispies,', 1),\n",
       " ('huda,', 1),\n",
       " ('jog', 1),\n",
       " ('bum', 1),\n",
       " (\"✨girl's\", 1),\n",
       " ('✨size', 1),\n",
       " ('✨new-paid', 1),\n",
       " ('✨search', 1),\n",
       " ('jkids', 1),\n",
       " ('deals!!✨', 1),\n",
       " ('tavas', 1),\n",
       " ('highlighters.', 1),\n",
       " ('luck', 1),\n",
       " ('pink(victoria', 1),\n",
       " ('secret)', 1),\n",
       " ('cas', 1),\n",
       " ('nwt......', 1),\n",
       " ('gray&white', 1),\n",
       " ('shirt..w/', 1),\n",
       " ('small-large.', 1),\n",
       " ('92', 1),\n",
       " ('lularoe!!', 1),\n",
       " ('45!', 1),\n",
       " ('(white', 1),\n",
       " ('raine', 1),\n",
       " ('cardi,', 1),\n",
       " ('cali', 1),\n",
       " ('summer!)', 1),\n",
       " ('frameless', 1),\n",
       " ('pictured)!!', 1),\n",
       " ('☆vs☆', 1),\n",
       " ('♡36dd', 1),\n",
       " ('♡baby', 1),\n",
       " ('♡good', 1),\n",
       " ('julia/jordan', 1),\n",
       " ('print!!', 1),\n",
       " ('pink,green', 1),\n",
       " ('boys/men', 1),\n",
       " ('pixar', 1),\n",
       " ('*cars*blanket', 1),\n",
       " ('33×44.', 1),\n",
       " ('matthew', 1),\n",
       " ('felder', 1),\n",
       " ('bowman', 1),\n",
       " (\"11/men's\", 1),\n",
       " ('cyphe', 1),\n",
       " ('ox.', 1),\n",
       " ('bali', 1),\n",
       " ...]"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from collections import Counter\n",
    "import operator\n",
    "sorted(Counter(' '.join(df[\"unprocessed_text\"]).lower().split()).items(), key=operator.itemgetter(1), reverse=False )"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Removing non english character "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "df[\"processed_text\"] = df[\"unprocessed_text\"].apply(lambda x: re.sub(\"[^a-zA-Z0-9\\s]+\", \" \", str(x) ))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Lemmatization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "from nltk.stem import WordNetLemmatizer \n",
    "ls = WordNetLemmatizer()\n",
    "df[\"lemmatized_text\"] = df[\"processed_text\"].apply(lambda x: \" \".join([ls.lemmatize(word) for word in x.split()]) )\n",
    "df = df[df[\"lemmatized_text\"]!=\"\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Functions to vectorize Text Columns and return a csr_matrix formatted vector and a list of feature names "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from scipy.sparse import csr_matrix, hstack\n",
    "\n",
    "def to_CountVector_csr(df):\n",
    "    n=4000\n",
    "    cv  = CountVectorizer(max_features=n,min_df=20, max_df=0.5, ngram_range=(0,2), stop_words=\"english\", binary=True)\n",
    "    count_vector_csr_matrix = csr_matrix(cv.fit_transform(df))\n",
    "    return count_vector_csr_matrix,list(cv.get_feature_names())\n",
    "        \n",
    "def to_TfidfVector_csr(df):\n",
    "    n = 3000\n",
    "    tfidf  = TfidfVectorizer(max_features = n, min_df=5, max_df=0.5, ngram_range=(1,2), stop_words=\"english\")\n",
    "    tfidf_vector_csr_matrix = csr_matrix(tfidf.fit_transform(df))\n",
    "    return tfidf_vector_csr_matrix, list(tfidf.get_feature_names())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "# This is a sparse martix containing the vectorized brands and Categories\n",
    "# column list will help keep track of the original features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "array_cat_brand, column_cat_brand = combine_category_and_brand_vectors(df[df[\"lemmatized_text\"]!=\"\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Word2Vec Vectorization"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Loading word2vec model\n",
    "# import gensim\n",
    "# word2vec_path = \"GoogleNews-vectors-negative300.bin.gz\"\n",
    "# word2vec = gensim.models.KeyedVectors.load_word2vec_format(word2vec_path, binary=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Tranforming Word2Vec\n",
    "\n",
    "# def word2vec_transformation():\n",
    "# #     w2v_processed_text = pd.DataFrame((df[\"processed_text\"].apply(lambda x: [word2vec[word] if word in word2vec else np.zeros(300) for word in x.split()])).apply(lambda x: sum(x)/len(x)).apply(lambda x: np.array(x)).values.tolist())\n",
    "#     w2v_lemmatized_text = pd.DataFrame((df[\"lemmatized_text\"].apply(lambda x: [word2vec[word] if word in word2vec else np.zeros(300) for word in x.split()])).apply(lambda x: sum(x)/len(x)).apply(lambda x: np.array(x)).values.tolist())\n",
    "# #     w2v_unprocessed_text = pd.DataFrame((df[\"unprocessed_text\"].apply(lambda x: [word2vec[word] if word in word2vec else np.zeros(300) for word in x.split()])).apply(lambda x: sum(x)/len(x)).apply(lambda x: np.array(x)).values.tolist())\n",
    "#     # Saving Transformed vectors as pickle\n",
    "# #     w2v_processed_text.to_pickle(\"w2v_processed_text_100000.pkl\")\n",
    "#     w2v_lemmatized_text.to_pickle(\"w2v_lemmatized_text_all.pkl\")\n",
    "# #     w2v_unprocessed_text.to_pickle(\"w2v_unprocessed_text_100000.pkl\")\n",
    "#     return w2v_lemmatized_text\n",
    "\n",
    "# pkl_file = word2vec_transformation()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### PCA retaining 95% of variance in data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "def pca_on_vector(df):\n",
    "    from sklearn.decomposition import PCA\n",
    "    pca = PCA(0.95)\n",
    "    return csr_matrix(pca.fit_transform(df))\n",
    "# Variance before PCA\n",
    "# variance2=pd.DataFrame(strat2Array).var()\n",
    "# variancePercent2=(variance2/sum(variance2.values))\n",
    "# variancePercent2=variancePercent2.sort_values(ascending=False)\n",
    "\n",
    "\n",
    "# plt.figure(figsize=(6, 6))\n",
    "# plt.plot(np.cumsum(pca.explained_variance_ratio_),color='#C41E3A')\n",
    "# plt.xlabel('Number of principal components')\n",
    "# plt.ylabel('Variance(%)')\n",
    "# plt.title('Scree plot of variance vs number of principal components')\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Target Columns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "y = df[\"price\"]\n",
    "y_log1 = np.log1p(y)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Scorer Function (not used)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "## Custom Scoring Functions\n",
    "from sklearn.metrics import mean_squared_error, make_scorer\n",
    "# Takes log1 values and spits out root_mean_squared_log1_error\n",
    "\n",
    "def root_mean_squared_log_error(y_true_log1, y_pred_log1): \n",
    "    return np.sqrt(mean_squared_error(y_true_log1, y_pred_log1)) \n",
    "scorer_root_mean_squared_log_error = make_scorer(root_mean_squared_log_error, greater_is_better=False)\n",
    "\n",
    "# This scorer first takes antilog-1 and the calculates mean percentage error\n",
    "def mean_absolute_percentage_error(y_true_log1, y_pred_log1): \n",
    "    y_true = np.expm1(y_true_log1)\n",
    "    y_pred = np.expm1(y_pred_log1)\n",
    "    return np.mean(np.abs((y_true - y_pred) / y_true+1)) \n",
    "scorer_mean_absolute_percentage_error = make_scorer(mean_absolute_percentage_error, greater_is_better=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature Engineered Arrays"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "from scipy.sparse import csr_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_numerical_csr = csr_matrix(df[[\"item_condition_id\", \"shipping\"]])\n",
    "columns_X_numerical = [\"item_condition_id\", \"shipping\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "array_tfidf_csr, column_names_tfidf = to_CountVector_csr(df[\"processed_text\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "strat1Array = hstack((X_numerical_csr, array_cat_brand, array_tfidf_csr)).tocsr()\n",
    "strat1columns = columns_X_numerical + column_cat_brand + column_names_tfidf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3118"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(strat1columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "# with open(\"strat1cols.txt\", 'w') as f:\n",
    "#     f.write(\"\\n\".join(strat1columns))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# with open(\"strat1cols.txt\", 'r') as f:\n",
    "#     mystring = f.read()\n",
    "# my_list = mystring.split(\"\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Comparing Models using Cross Validation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Takes Featurized vectors, num_CVs and title as input and plots Model Comparison \n",
    "def cross_validation(X, y, cvs, strategy_name):\n",
    "    from sklearn.neighbors import KNeighborsRegressor\n",
    "    from sklearn.linear_model import LinearRegression, Ridge, Lasso\n",
    "    from sklearn.model_selection import cross_val_score\n",
    "    import time\n",
    "    import seaborn as sns\n",
    "    import xgboost as xgb\n",
    "    import lightgbm as lgb\n",
    "    from sklearn.ensemble import RandomForestRegressor\n",
    "   \n",
    "    # Make a list of models that you would like to train on(Ref: https://github.com/curious-attempt-bunny/ml-katas/blob/master/abalone-kata-4.ipynb)\n",
    "    models = []\n",
    "    models.append(('Linear Regression', LinearRegression()))\n",
    "    models.append(('Lasso', Lasso(alpha=0.001)))\n",
    "    models.append(('Ridge', Ridge(alpha = 20)))\n",
    "    models.append(('XGB Regressor', xgb.XGBRegressor(random_state=42, silent=True)))\n",
    "    models.append(('Light GBM', lgb.LGBMRegressor()))\n",
    "    models.append(('Random Forest', RandomForestRegressor(max_depth=40)))\n",
    "    models.append(('KNN', KNeighborsRegressor(n_neighbors=8, algorithm=\"kd_tree\", weights=\"distance\")))\n",
    "\n",
    "    results = []\n",
    "    names = []\n",
    "    start_time=time.time()\n",
    "    for name, model in models:\n",
    "        cv_results = cross_val_score(model, X, y, cv=cvs, scoring=\"neg_mean_squared_error\")\n",
    "        results.append(np.abs(cv_results))\n",
    "        names.append(name)\n",
    "        msg = \"%s: %3.2f (%3.2f)\" % (name, -cv_results.mean(), cv_results.std())\n",
    "        print(msg)\n",
    "    print(\"Time Elapsed:\",time.time()-start_time)\n",
    "   \n",
    "    plt.figure(figsize=(10,7))\n",
    "    plt.title('Comparision of Models using 8 Fold Cross Validation')\n",
    "    sns.boxplot(x=names, y=results, palette=\"Set3\")\n",
    "    plt.xticks(list(range(len(models)+1)),names)\n",
    "    plt.xlabel(\"Model Names\")\n",
    "    plt.ylabel(\"Cross Val Scores (RMSLE)\")\n",
    "    plt.grid(True)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model Stats\n",
      "Linear Regression: 0.44 (0.02)\n",
      "Lasso: 0.36 (0.02)\n",
      "Ridge: 0.34 (0.02)\n",
      "XGB Regressor: 0.39 (0.02)\n",
      "Light GBM: 0.34 (0.02)\n",
      "Random Forest: 0.39 (0.02)\n",
      "KNN: 0.46 (0.02)\n",
      "Time Elapsed: 79.90827202796936\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x504 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(\"Model Stats\")\n",
    "cross_validation(strat1Array, y_log1, 8, \"Strategy\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Default Model and it's performance"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "RMS Log error of default model: 0.740\n",
      "R squared: -0.001\n",
      "y_pred: mean of prices = 18.36 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Baseline Model Stats takes log prices as input\n",
    "from sklearn.model_selection import train_test_split as tts\n",
    "from sklearn.metrics import r2_score, mean_squared_error\n",
    "def baseline_model(X,y):\n",
    "    X_train, X_test, y_train, y_test = tts(X, y, random_state=42,test_size=0.3)\n",
    "    pred = y_train.mean()\n",
    "    y_pred = [pred for i in range(len(y_test))]\n",
    "    print('RMS Log error of default model: %1.3f'%(np.sqrt(mean_squared_error(y_test, y_pred))))\n",
    "    print(\"R squared: %1.3f\" %(r2_score(y_test, y_pred)))\n",
    "    print(\"y_pred: mean of prices = %1.2f $\"  %np.expm1(pred))\n",
    "    \n",
    "    sns.scatterplot(np.expm1(y_test), np.expm1(y_pred))\n",
    "    plt.xlabel('True Price')\n",
    "    plt.ylabel(\"Predicted Price\")\n",
    "    plt.title(\"Default model\")\n",
    "    plt.grid(True)\n",
    "    plt.show()\n",
    "    \n",
    "    \n",
    "baseline_model(strat1Array, y_log1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Plotting Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scatter Plot of results\n",
    "import seaborn as sns\n",
    "def scatter_plot_of_results(y_test, y_pred, label, linear=True):\n",
    "    sns.scatterplot(y_test, y_pred)\n",
    "    \n",
    "#     sns.regplot(x=y_test,y=y_pred)\n",
    "    plt.axhline(y_test.mean(),color='black')\n",
    "    plt.title(label)\n",
    "    plt.xlim(0,300)\n",
    "    plt.ylim(0,100)\n",
    "    plt.xlabel('True Prices')\n",
    "    plt.ylabel(\"Predicted Prices\")\n",
    "    plt.grid(True)\n",
    "    plt.show()\n",
    "    \n",
    "    print(\"R_2 score: %1.3f\" %(r2_score(np.log1p(y_test), np.log1p(y_pred))))\n",
    "    print(\"RMSLE: %1.3f\" %(np.sqrt(mean_squared_error(y_test, y_pred))))\n",
    "    print(\"Prediction of default model: %1.2f $\"  %y_test.mean())\n",
    "    \n",
    "    \n",
    "def display_model_predictions(model, X_train, y_train, X_test, y_test, idx_train, idx_test, title):\n",
    "    model.fit(X_train,y_train)\n",
    "    y_pred = model.predict(X_test)\n",
    "    scatter_plot_of_results(np.expm1(y_test), np.expm1(y_pred), title)\n",
    "    result[title]=np.round(np.expm1(y_pred),2)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split as tts\n",
    "X=strat1Array; y=y_log1; indices = np.arange(X.shape[0])\n",
    "X_train, X_test, y_train, y_test, idx_train, idx_test = tts(X, y, indices, random_state=42,test_size=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.165\n",
      "RMSLE: 38.117\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAY4AAAEWCAYAAABxMXBSAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy9eXxU5b34//7MlkwWSAgB0cS1GBsRhCiL9N6LpdVaUWtBbQVcUAGp0m/rgv213OqlfV1xqbfeiiBtVXCpKPVq7deltaL3i+ISUWsRpIrKnhATyDKZ7Ty/P86c4cwWZpJMZhKe9+uVVzJnzvKcmZzn83x2UUqh0Wg0Gk26OHI9AI1Go9H0L7Tg0Gg0Gk1GaMGh0Wg0mozQgkOj0Wg0GaEFh0aj0WgyQgsOjUaj0WSEFhyavEZEpojIjlyPQ6PRHEQLDk23EJHPRMQnIm0iskdEHhKRklyPKxPi7sH6ObIPr3+siCgRcdm2XSEi4chYDojI+yIyra/GpNGkgxYcmp5wnlKqBDgVGAv8JMfj6Q7nKaVKbD+7MjnYPun3Im9EPtcyYBnwBxEp6+2LZGnsmVxfRETPQf0Q/aVpeoxSag/wIqYAQUTOFZGNkRXzdhG51drXtsq+XES+EJF9IvJT2/veiPbSLCKbgNPt1xKRr4rIOhFpEZF/iMj5tvceEpFlIvJ8ZMW+XkSOEJH/ipxvs4iMTeeeROT8yPlbItf7qu29z0RkkYh8ALSLiEtEjhSRtSLSKCLbRGShbf/xIvJO5PPYKyK/irz1WuR3S2S8k+I+VwNYDRQDI23nmygir0fG9r6ITLG9d5yIvCYirSLyVxG5T0QeifvsrxKRL4C/pXG+K0Tk08j5tonIzMj2r4jIqyKyP/IdPmE75gwReTvy3tsicobtvXUi8ksRWQ90AMen831o8gyllP7RPxn/AJ8B34j8XQX8Hfh15PUU4BTMhcloYC/wnch7xwIKWAl4gTGAH/hq5P3bgf8FhgDVwIfAjsh7buCfwP8HeICvA61ATeT9h4B9QB1QiDkxbgMuA5zAL4BXkt1D3L2dCLQD34xc8+bIdT22496LjM8buc964N8j4zoe+BQ4O7L/G8DsyN8lwMS4z8Jlu/YVwP+L/O0EfgAEgGGRbUcBTcC3I9f9ZuR1pe1ad0XG8TXgAPBI3PVWYQojb1fni+xzwPb5jgBOjvz9OPDTyDGFwNci24cAzcBswAV8P/K6IvL+OuAL4OTI++5c/y/rn248/7kegP7pnz+RybMtMnEr4GWgLMW+/wXcE/nbmryqbO+/BXwv8venwLds783loOD4F2AP4LC9/zhwa+Tvh4CVtveuBz6yvT4FaElyDy2Rn/+JbF8MrLHt5wB2AlNsx82xvT8B+CLunn8CPBj5+zXgNmBo3D7WZxEvOEKR8QQBH3Cx7f1FwOq487wIXA4cHTm2yPbeIyQKjuPTPF9xZBzTAW/cPquAB+zfY2T7bOCtuG1vAFdE/l4H/Eeu/3/1T89+tKlK0xO+o5QqxdQwTgKGAojIBBF5JWK22Q/Mt96zscf2dwfmShzgSGC77b3PbX8fCWxXpgnH/v5Rttd7bX/7kryOd+B/RylVFvn5ju060etGrrc97jr2MR4DHBkx9bSISAumVjQ88v5VmFrM5ojp5lDO7g1KqTKgHHgWU2Dar3VR3LW+hqkNHAl8qZTqSDHOVGNPej6lVDtwCeb3t1tE/iwiJ0WOuxkQ4K2ISW9OZHvMZxch/jtKNiZNP0ILDk2PUUq9irnavyuy6THMCa9aKTUYWI45yaTDbkwTkMXRtr93AdVxDtWjMbWB3mQX5oQKmE7cyJjs17GXld4ObLMJoDKlVKlS6tsASqmtSqnvA8OApcBTIlIcd44ElFJtwAJgts03sx1TQ7Bfq1gpdTvmZzdERIpsp6kmkfixpzofSqkXlVLfxBRMmzFNjCil9iilrlFKHQnMA5aJyFfiP7sI8d+RLsndz9GCQ9Nb/BfwTRE5FSjFXPl2ish44NIMzrMG+ImIlItIFaa5yeJNTN/DzSLijjhxzwP+0Ct3EDuGc0Vkqoi4gRsw/TCvp9j/LeBAxGHuFRGniIwSkdMBRGSWiFRGNJeWyDFhoBEw6MJBrJRqAn6L6T8B0/R0noicHblOoZi5LlVKqc+Bd4BbRcQTcbafd4h7TXk+ERkeCRIojtx/W2TciMhFke8HTB+Girz3f4ETReTSSNDAJUAt8NwhxqHpR2jBoekVlFKNmHbvxZir5P8QkVbMCW9NBqe6DdO0sQ14CTOqyLpGADgfOAfTCb4MuEwptbk37sF2nS3ALOC/I9c5DzNsN5Bi/3Bkn1Mj496HOdkPjuzyLeAfItIG/BrTn9MZMSn9ElgfMRNNTDGk/wK+LSKjlVLbgQswTWGNmBrDTRx8lmcCkzAd3L8AnsCc9FPda1fnc2AKzV3Al8C/YX63YEa7vRm5p2eBHyqltkUE3bTIcU2YJq1pSql9qcag6X+IUlpr1GgGKpEw2c1KqZ/neiyagYPWODSaAYSInC4iJ4iIQ0S+halN/E+ux6UZWGRNcIjI70WkQUQ+tG0bIiJ/EZGtkd/lke0iIveKyD9F5AMRGZetcWk0A5wjMENe24B7gWuVUhtzOiLNgCObGsdDmLZdO7cALyulRmLG/d8S2X4OZmbsSMy4/fuzOC6NZsCilPqTUqpaKVWklDpRKfVgrsekGXhkTXAopV7DdKjZuQB4OPL3w8B3bNtXKZMNQJmIjMjW2DQajUbTffq6yNlwpdRuAKXUbhEZFtl+FLFJQTsi23bHn0BE5mJqJRQWFtYdffTR8bsMGAzDwOEYuG6ofLg/e2hIuokm6ZAP95ZN9P31bz7++ON9SqnK7h6f0+qYNpI9s0nDvZRSD2CWOqCmpkZt2bIlm+PKKevWrWPKlCm5HkbWyOX9GYZiy95Wrln1DjuafVSVe1l52WnUDC/F4ei5CNHfXf9moN+fiMRn92dEX4vUvZYJKvK7IbJ9B7EZrlWYseMaTVZoag9EhQbAjmYf16x6h6b2pKkaGo3GRl8Ljmcxi6cR+f2MbftlkeiqicB+y6Sl0WSDQCgcFRoWO5p9BELhHI1Io+k/ZDMc93HMqpg1IrJDRK7CLJn9TRHZilm++fbI7v8XsyrqPzFr4SxIckqNptfwuJxUlXtjtlWVe/G4nDkakUbTf8iajyNS1C0ZU5PsqzD7Dhz2GIaiqT1AIBQmZCgMQ/WKzV0TS0Wxh5WXnZbg46go9uR6aBpN3pMvznENiQ7bn5xqcNTe1l5z2GoO4nAINcNLeXrBZAKhMB6Xk4pij/6cNZo0GLjxZv2QeIdtIGxoh20WcTiEytICjiovorK0QAsNjSZNtODII7TDVqPR9Ae04MgjtMNWo9H0B7TgyCMsh60lPDxOh3bYajSavEM7x/OIeIftRxvf1I5xjUaTd2jBkWdYDluArQ7RQkOj0eQd2lSl0Wg0mozQgkOj0Wg0GaEFh0aj0WgyQgsOjUaj0WSEFhwajUajyQgtODQajUaTEVpwaDQajSYjtODQaDQaTUZowaHRaDSajNCCQ6PRaDQZoQWHRqPRaDJCCw6NRqPRZIQWHBqNRqPJCC04NBqNRpMRWnBoNBqNJiN0P44BgGEomtoDBEJhPC4nFcUe3cdDo9FkDS04+jmGodiyt5VrVr3DjmYfVeVeVl52mu4cqNFosoY2VfVzmtoDUaEBsKPZxzWr3qGpPZDjkWk0moGK1jh6kVyYjAKhcFRoWOxo9hEIhbN6XY1Gc/iiBUcvkYnJqDcFjMflpKrcGyM8qsq9eFzOHt2PRqPRpEKbqnqJdE1GloC5cNl6Ji99hQuXrWfL3lYMQ3XruhXFHlZedhpV5V6AqMCqKPb07IY0Go0mBVrj6CXSNRmlEjBPL5hMZWlBxtd1OISa4aU8vWCyjqrSaDR9ghYcvUS6JqNs+CQcDumW0DkUOsxXo9EkQ5uq0sAwFI2tfnY2d9DY6k9qVkrXZGQJGDv56JPobZOaRqMZOGiN4xCk6/RO12RkCZj48+WbT6K3TWoajWbgoAXHIchkAk3HZNRffBI6zFej0aRCC45D0J98Er2JDvPVaDSp0D6OQ9BffBK9jQ7z1Wg0qciJxiEiPwKuBhTwd+BKYATwB2AI8C4wWymV87oZ/cUn0dv0F5OaRqPpe/pccIjIUcBCoFYp5RORNcD3gG8D9yil/iAiy4GrgPv7enzxHM4TaCqTmg7T1WgOb3Ll43ABXhEJAkXAbuDrwKWR9x8GbiUPBAf0D59EX6Gr8Wo0GlGq7+PyReSHwC8BH/AS8ENgg1LqK5H3q4HnlVKjkhw7F5gLUFlZWbdmzZo+G3df09bWRklJSa6HEUPIUHzS0EYgbES3eZwOThhWgitDwdHd+wsZCqUUIpLxNfuKfPzuehN9f/2bM888s14pdVp3j8+FqaocuAA4DmgBngTOSbJrUommlHoAeACgpqZGTZkyJTsDzQPWrVtHvt3fzuYOrl76CvFxFesXTeCo8qKMzpXp/SXXduryUtvJx++uN9H3d3iTi6iqbwDblFKNSqkg8EfgDKBMRCxBVgXsysHY8p74LPZQyDhkVntvkssoM917RKPJD3IhOL4AJopIkYgIMBXYBLwCzIjscznwTA7GltfElwH56dMfsLmPy4LkMky3vyUl9qVA12j6kj43VSml3hSRpzBDbkPARkzT05+BP4jILyLbftfXY8t34lfc0+uqmf9IfZ+WBclllFl/SUo0DEVn0ODCZet1AIFmQJKTBECl1M+VUicppUYppWYrpfxKqU+VUuOVUl9RSl2klPLnYmz5TPyKu8zrzskK3IoyO6q8iMrSgj6bDPtLUmJTe4DPm9q1SU0zYNElR/oR8SvuFl+wRyvw/paP0V9yagKhcCTq7OC6LJ9NahpNpuiSI31AOmXZ0yF+xb22fjvLZ9V1awXeX8um50rbyQSPy4nHGfto5aNJTaPpLlrjyDK9mTCXbMVd7nV3awWuy6Znj4piD8dUFFNVzmFVpkZz+KAFR5bJRqvY+OO6c57+FqHUE/raJOdwCIVuB08vmJTXJjWNprtowZFl8nWC7i8RSj0llyVStOamGahoH0eWydey7P0lQqmn6KRBjab30RpHN8jE9JGvZdn7S4RST8lXjU+j6c9owZEhmZo+8nmCPhyq/h4uJjmNpi/RpqoM6Y7poz+EkA5UDheT3OFEb4W3a7qP1jgyRJs++hf5rPFpMkf3g8kPtMaRIfnq7NakRmt8Awcd7JAfaMGRJpZ6HAiFeezqCZxVOwzQpg+Npi/RGn9+oE1VaZBMPV4xu44lF4zC4XBo04dG00foYIf8QGscaZBMPZ63uh6Hw6FNH93A0t6CYUM7NzUZoYMd8gOtcaSBVo97D7v2dklVK7ctW6+dm5q00cEO+YHWONJAO8R7D+3c1PQUHeyQe7TgSAOtHvceWnvTaPo/2lSVBlo97j20c1Oj6f9ojSNN+qN6nI8Ztlp702j6P1rjGKDka4atXXt77631PH3+ZK29aTT9DK1xDFDy2QltaW9upw5n1mj6I4cUHCLyQxEZJCa/E5F3ReSsvhhcfyEfTULaCa3RaLJFOhrHHKXUAeAsoBK4Erg9q6PqR1gmoQuXrWfy0le4cNl6tuxtzbnw0CHEGo0mW6QjOCw7wreBB5VS79u2Hfbkq0lIO6F7l3zUKjWaXJGOc7xeRF4CjgN+IiKlgJHdYfUf8tUklC8hxJl0S8xX8jXQQKPJFekIjquAU4FPlVIdIlKBaa7SkN95Cbnu8DdQJtxUWuXTCyYP+A6KGk0y0jFVKaAWWBh5XQwUZm1E/QxtEkpNvprxMiVftcrDHW0+zB3paBzLME1TXwf+A2gF1gKnZ3Fc/YZ8MQnlI30x4faFKSyftcrDlYGizfZX0tE4JiilfgB0AiilmgG9nLbRH7PK+4JsR3b1VUSb1irzj4GizfZX0tE4giLixDRZISKVaOd4vyDXjmlrwo1fFfbWhNtXvgetVeYf2nyYW9IRHPcCTwPDROSXwAzgZ1kdVT8n1xO2NYZcq/LZnnD7cvLIdaDB4Uyy50mbD3PLIQWHUupREakHpmLmb3xHKfVR1kfWT8mHCRvyJxIomxOunjwGPqmep5GVJVnVZjVdk07JkYnATqXUfUqp3wA7RGRC9ofWP8kX22s2V+PZimbJ9Lza9zDwSfU8NfuCUW12/aIzeXrBZO0Y70PSMVXdD4yzvW5Psk0ToTcmbEs1t3pyd8e8k63VeLY0qu6cV/seBj5dPU/afJg70io5opSKLv2UUgY9LMcuImUi8pSIbBaRj0RkkogMEZG/iMjWyO/ynlwjV/Q0ksgeKbR5T2u3I4WytRrPlkbV3fPqiLaBja65lp+kIzg+FZGFIuKO/PwQ+LSH1/018IJS6iRgDPARcAvwslJqJPBy5HW/o6cTdm9NzPbVeG+q8tkygekoGU0ytDkyP0lHc5iPGVn1M8yQ3JeBud29oIgMAv4VuAJAKRUAAiJyATAlstvDwDpgUXevkyt6aj7pzQk0G6p8tkxg2tGtSYY2R+YnYrNC9c0FRU4FHgA2YWob9cAPMR3wZbb9mpVSCeYqEZlLRHBVVlbWrVmzpk/G3VeEDMUnDW0EwgbDvbDXBx6ngxOGleDKk4elM2jweVM7gbCBx+ngmIpiCt2Z9wRra2ujpKSk18+bD8Tf20BD31//5swzz6xXSp3W3eNTCg4RuVkpdYeI/DeR5D87SqmFSQ479AVFTgM2AJOVUm+KyK+BA8D16QgOOzU1NWrLli3dGUbeYncSX1LVyhM7SrMezptp3klv5amsW7eOKVOm9Pp584H4exto6Pvr34hIjwRHV6YqK1fjne6ePAU7gB1KqTcjr5/C9GfsFZERSqndIjICaOjl62aF3p7s+rond3ejmbIRzaKjZDSa/kFKwaGU+lOk1MgopdRNvXVBpdQeEdkuIjVKqS2YiYWbIj+XY3YXvBx4preumS2yFZoa35M7m+RLoqBGo+k/dOkcV0qFRaQuC9e9HnhURDyYEVpXYkZ4rRGRq4AvgIuycN1eJZNJN1/NMDqa6fAjX/8XNf2HdKKqNorIs8CTmMl/ACil/tjdiyql3gOS2demdvecuSB+0h1bXcb8KSfQEQjR2Er0gcyXMiTJ0NFMhxf5/L+o6T+kE7IyBGjC7MdxXuRnWjYH1V+wJyeNrS7jxrNrWPLcJv7tznUxiXv5UoYkGTpO/vAin/8XNf2HdDSOm5RS+7I+kn6IvWz4/CknsGjtB0nNVj0xBzW2+rNqUshWnLw2h+Qn2jSp6Q1SCg4ROQ/4PWY/DgO4WCn1ep+NrB9gn3Q7AqGUD2R3zEGGoegMGly4bH1Kk0JvTc69Hc2kzSH5izZNanqDrkxVvwT+RSl1JDAd+M++GVL/wpp0izyulDV1ujIHpaoI29Qe4POm9pQmhb7qftcdtDkkf9GmSU1v0JWpKqSU2gwQSdQr7aMx9RvsK363y8GqOeO57PdvJfQHSGUOAlKuzAOhMIGwgV22200K6UR05cpcpM0h+Ysu4aHpDboSHMNE5MepXiulfpW9YeU/ycwxy2fVcc/Fp2IoRUcgTIHr4KSfzBzU2OpPOfl7PU6OrSjmibljafEFWb7uExrb/FGTwqEm51yai7Q5pG/o7sJAJ1pqekpXgmMlUNrF6wFDdx7AZCv++Y/Us3haLfNW1wPmZNlVIl2qyd8wDPa1BdnV4uP6BzZQVe7lzhmjGT6oMKqpHGpyzmViX7Z7jWu0H0mTW7rKHL+tLweSK7r7AKaa9Mu87pjXXZlnPC4nZ9UOY3pdNWVeNy2+IGvrtxNWRGpVmaaqHc0+bnrqA9bMnUhTe4CKYk/C5HxW7TB+dm4tgVA44isxcmYucjiEkZUlrJk3iVDYwOV0MKxE98roTXTGvyaX9Kgh00Ag0wfQ0k6ApCv+Fl8w5nVX5plyr5uFU09k/iP1MeYup5jjoOrgvjuafWxv9nHDk+9HBZtlqzYMg33tAS797ZvR86yYXcdZtcN4adPBkl99ZS4yDMXWxja9Gs4i2o+kySX9s2Z1L5LJA2iPZLrusY3cOWN0THTK8ll1rK3fHn19qMipZl8wKjSs685/pJ6wImmEVosvGBOhZNmqHQ4H81bHnmfe6np+dm5tr0XPZNIPXEdVZZ9cdcaz/g+stsb5EMWn6XsOe40jla9AxCwVYl8h2yfEHc0+7nhhC0suGMUJw0rwup2Ue9388sLR/Py89COnkgktpRQrLzuN1159NTqepdNHc9eLW5KWNUl1HqdDeiV6JlNznl4NZ59c+JHiS/7ftmy91iQPU7pKAPxxqvegf0RVpeP0TvYA3n3RGG599kN+9M2amIcifkLcuL2FKx96m9dumgJkHjnVlYO7ZriXncNKeO3mCXzS0MZdL5p9R248uyaaoR6dLEo8Kc/TG/buTM15Oqoq++QirDbm/6BK+1UOZ7oyVZVGfk4DrgWOivzMB2qzP7SekUmCXIHLwZILRvHE3IksuWAUBW4Hja2BBPNKKvPAR3vM63y05wChkBHzfler766SsRwOweUQqsq8HDG4kMY2f8qyJi6HZDWpy7qHsdVlrJhdxxNzJ7J4Wi2GYSTdXyeZ9Q3WQuWo8qKIyTK7q36tSWosDhlVJSIvAeOUUq2R17diVsrNa9JdJTe1B6JJexZV5d5oWK1hGNF6UV6PM0E7sUxIll/hsasnUFVeFH2Iu1p9p7NqTKesiS8Qzurq04r+uvyM42K0nRWz66goLqDZF0y4bnfGo+tb5Tdak9RYpOPjOBqwezUDwLFZGU0vku7qqKuw2rNqh7GvPRB1PFeVe1k1Zzx/XHAGnYEwH+1p5a4Xt7Bxe0v0uIZWP16PKyqckpnCVsyuozwStptOMpa1T2Nr8kguSwhly1xQUezhZ+fWRqO2rHv99V8/5offODHm87HbvDMZj85LyH/s/8vQqjXJw5h0oqpWA2+JyK0i8nPgTWBVdofVc9KNOkm1X0cgzM/OrU2IVrrs928hCF6PiyXPbYoKDes4a8VsYeU0PHb1BJ6aP4nF02r59V8/ZmtjW5cRKYahCBkqJoopVyYgh0NwOiT6OVgmq5u/dRINB/xUlpgCoifRUzoSK/+xa5InHWH+ji+6mW7knaZ/c0iNQyn1SxF5HviXyKYrlVIbszusnpNu1ElSjWBWHSPKCvEFUmstIwabmoN9tb10+mgefn0bY48+JeaYZl8wZrUOsGl3a5e5Ilv2tvJJQxtXL30lZvWdLZPUocxEloCtLClIcNBb5rqN21u6bfPW9vP+Qaq2xlpjPLxINxy3CDiglHpQRCpF5Dil1LZsDqynpGtn72q/RsMfNQ1ZYbAVxR5EzHMcVVbIqjnj+bI9QFN7gIdf38aVk4/DFXeNTCdFa/Vtzxy3+2d62ySVzkNvCdg9+zsTHPSL1n4Q9Ql11+at7ef9G53JfnhxSFNVxDy1CPhJZJMbeCSbg+otDhV1YqnWu/eb/+wjBntj9rMmy7Nqh0W7+81Y/gYXr3iDLXtb8QcNbljzPk3tAcq8bqbXVXPHC1vwBWIFQqbJWr2x+u7thD1LwJ4wrDilT6gnpjMdidW/yQeNUZvK+o50NI4LgbHAuwBKqV0DocR6Oqtsa7K89fxRXLzijYSJdc28STS2+aNFDSG5QCj3ulk+qy6htEi5ra6VnYOCprXL8/bk3uyk+9A7HILX7UqqGVgFHXtiOhs+qIAn5k4krKDQ7WBosa5v1V/ItcaoTWV9SzrO8YBSSgEKQESKszukviFdZ6zDISiluszwPtQqudkX5N6XP2bxtFqeXnAGD15xOiLQ0JZ8VWStvj1OR5fn7em9WWSiEaXSDOK1NUh/BWg99Of/xsy5uXTlBpratFO8P5FrjVEHV/Qt6Wgca0RkBVAmItcAc4DfZndY2ScT1fpQGd6H8qMEQmFe2tRAY2uAG8+u4frHN3a5KrI0nZ3DSli/aMIhneDxju1UlXHtOSn2c6YKJCj3upPun47vKNkKcMXsOoYWe3A4HDETiraP939y3SAqH0xlhxPpRFXdJSLfBA4ANcC/K6X+kvWRZZlMVOuuIrSS5SvET+Rej3mtVJnfySZIK3P8qPKiLu8j1QQdXxk3WU6KXWjFP/TlXneXFW4PNaEnEwbzVpv9SpY8t4mVl50W3Vc/9AODXDaIyrWp7HAjHef4UqXUX5RSNymlblRK/UVElvbF4LJJMtV61ZzxKFSCacU+sa5fdGZC/LqdZKVO9h7ws2rOeCqKPb0+QaaaoOMr4y6ednJCTopdlY8PJGj2BXuk+neVWGmdKxT5fHNV6VUzcMi1qexwIx1T1Tcxo6rsnJNkW97RVW5C/Crb63Gy94Cfy5a9nnSFne5qKpXZ5Y8LzqDQnX4l3nTvKZ3KuF6PkwO+5OVKUgmtnmoBqVaAwbARPZdS5kN+qOx6jeZQ9IapzP5shQyV9jN5OJJS4xCRa0Xk78BJIvKB7Wcb8Pe+G2L3SKfIoX2VHTboFedaqgk3GDI4YlBhwqpo6fTR3Prsh2zZ20ooZMQ4k9O5p7Chkq7W3S5HzL1t29eecr9k9FQLSLYCvHPGaEoKXYytLosKTMg8u16HXWqS0ZOij/HP1icNbSmLomq61jgeA54H/hO4xba9VSn1ZVZH1Qtk6nDtLTt7OkUN18ybxK4WH03tgZhaV/F1n342TvFlu58yr7lyir+nypIC9h7wJ4T63jljdEwSYiAU5t6Xt7J0+uiYjO/7Z46jrTPE0GIV/czsPo6e9HtwOIThgwpYcsEoijxOWnxB7nhhC41tfpZcMIpjKor49O9vs7O5A4/LidNBWtn1OuxSkw3in61A2NABGl3QVXXc/cB+Efk18KWtOm6piExQSr3ZV4PMFMNQBEJh7r5oDC2+IMvXfZKyHIalnoaV4sErTufel7dGJ/Kqci9ejzNpZBFAKGTQ0GZ2Q3NH+mofqtSJFd47Y/kbMeOYXled4IP4vMmgYPt+jhhcmND4aWx1GTeeXcOP1rxHZYk5QR9dUcTuFrPB1G8uHQuRwGmPy0ljm5+7XtzC4mm1lHnddATCtPlDXOGbXPMAACAASURBVPvouzx73WT2HvAnjHlkZUmPVH9fIMyVD72dsP0rw0po84diSqqsmFVHZUlBjOBI9n3pCCxNNtABGpmRjo/jfmCc7XV7km15Q7IVqVVLqbHNH2NqSbbvnTNGR1fGq+aMTzqh1gwvxTAUm/e2JiT1nZRGPalkWkkyx3kgbFDkcSZt/GSP0NrR7OPKh96OloOPv0+7MLPKgiydbt7njmazLHs2JuNU2pfTIQklVeY9Us+SC0bFCJpkpjH9gGuygY7Kyox0EgAlkgAIgFLKII9bziZbkS5a+wELp45MMLUk2/empz7gV5eM4bGrJ1BS4Erp92ho8yftF97Q5k9qa7Xb5RWKVXPGx9j/h5UWJPgUPE5HtM94fOMnKzrJzo5mX3Qf+33aTWSWD8EykVWVewmnSHDs6WRsZczb73P5rDqcQtLrHTe0+JBRMToCS5MN4n1yHqdDR2V1QToC4FMRWYipZQAsAD7N3pB6RqoV6QnDSqgq88as/FPt23DAzw1Pvs/j10xIOaGGjOSTbSic2BUvlV3+2esm4wuk9ilUlRssX/dJgo/E0maSrZCOKvMyfFAhQIKJ7YhBhez3Bfk/T7wXM45U0V49nYztGfNlXjctkde3nj8qaUmVogLnIU1juei1rRn4xD9bH218U/vNuiAdwTEfuBf4GWbZkZeBudkcVE9IpXJ63c6Ef4JU+1qrfIdI0vfBXJEke8/lPKjEWf4TXzDEnv2dURu+pbk8c90Z0ZDaZh+MrCzhjwvOoMMfZtu+dj79+9s0tpkrH6eDqCPZmiRXzj6Na1YfnEDvuXgMvmAYw1Apk/eSmdGAHk/GyUKfrYx5eyIiwJILzOu99uqr0c9t5WWnpVWbKtcZypqBiz3kfmskBF+TnHQyxxuA7/XBWHqFTFakyfa1/CFV5V4cDhIikZZOH02bP0RpoSshmmnZzHGEDCMawpfK1wJw87dq2N3ijznemtwphuICF6GdXv74rTNo6wxx/m/WJziu3S6JRi11BMK4XQ5uf/4jbj1/VJc+i2R+i55Mxqk0qooST1Lh6nA4MiqpEk8uM5Q1Gk0XgkNEblZK3SEi/02kwKEdpdTCrI6sm2SyIrX2ta/yLSe6ucp38PDr22JMLQ+/vo1bzvkqF6/YEI1mOnZoEUrB8nWf8PqnTTy9YDKQmBeyaO0H3P7dUxAROoMGD66PNePc85ct/PLC0dHJ3e10IEhMT3RLCKyZN4krHnw7YVJePK2WYDh5raqufBY9mYxTRTo9e93kLku1pFNSRaPR5B9daRwfRX6/0xcD6U3SnQTt4bQel4OxRw/mN5eOjTHh/OibNTET332XjuP25z9KGs20pn4HQHSCTjZ5Vw0pYltjOyMGF3D5GcclaDOGEesjiffDWA2lgmGDB684nc5gmF37O6Mhx1XlXlyO5Ca2ZD4Lw1C0+AL4AmHCSlHodmZczjyVr8gXCGuzkkYzAOkqj+NPkd8PZ+PCIuLEFEo7lVLTROQ44A/AEMzeH7OVUj2uiZyq7EgoZKQMp3XZsqnjE/YcAo2tAVbMrotqCsvXfUJZpDyGNUE7HfDgFadHk9+Wr/uExjY/O77sAKCk0M3Vq+oTNJI18ybFjN/uh7FyN+KFzdr67dx4dg2vbdmLUnDbn/6RYGJLZq4zDMVnTe3sPdDJTU99kGgyIzEpsNkXTPgsD5X0OBDMSodqravRHE50Zar6E0lMVBZKqfN7eO0fYmo1gyKvlwL3KKX+ICLLgas4GMnVLbrKMk4VTvvE3ImMGGw6wK2Jwu1yMKTYw/954j3unDGam79VEzPR3jljNIZS0RpLbifsbOlk8TMfxuxTUuAirBTXPbaRuy8ak3SVbot8BmL9MMmq61rmr0DIYPYZx/G9Bzawo9lHY2uAxdNqqSj2cGSZlyMGFSZMdE3tAT5v6oiO0zqnZWaKz2FZPquOe1/+mJc2NcR8ltmOdOrLSTvZtSDRX6Wz1TWHM12Zqu6K/P4ucAQH28V+H/isJxcVkSrgXOCXwI/FLFr0deDSyC4PA7fSQ8HRVZYxqBj/gmXq8YcMPmtqxx8yoseeVTuMn51by+PXTMAhwm1/+kdUA5g/5QTcTgdHDC7kyXkTCRmKjV/s5/G3Po85/4PrTd+I5a9o8QXTMifZfTYdgeSFCkeUebn892/FCKON21uinQnXLzozZee/Io8zpZkp/rOb/4hZFv2lTQ0JDvdsmaR6q8RIOsInFDLYtd9HQ6ufpvYAa+u386Nv1jB8UIHOVtdobEj8CjdhB5HXlFL/eqhtGV1U5CnMGlilwI3AFcAGpdRXIu9XA88rpUYlOXYukXDgysrKujVr1qS8TjBssHlPa8L2k44oJWwoPm/qIBA2oqG1+9oCHFlWiKHgs33tkcxtF0cMMsNo7fu2dAQpK/aw48uD5zimooi9B/wcMbgQAUKGImQomtsDDC3x4HE5ouMZUuRhSImHLyJjGFToZkSZeRwIAvg62igtLbXdj+LTxjYCtlwRj9PBkWVePmtq55iKYna3+BLeP76yGBGJ1q8KGSqi2YjpI0lyzHGVxWxJ8tkdX1nCp41tMZ+l25lOHmkibW1tlJSUdLlPyFB80pB4zycMK4mpx3UoOoMGnze1276rYgrdjrh9win/J1L9H6W693TurT+j769/c+aZZ9YrpU479J7JSSePo1JEjldKfQoQ8UVUdveCIjINaFBK1YvIFGtzkl2TSjSl1APAAwA1NTVqypQpyXYDzAS425atT1jVP/HtiVzywAZ2NDuwkueryh2smvOv3LDmfW49/2Tm/XU94GDF7DqWPLcpYd+H5/wLl//+rbjtwn9/fyJupzNiBuuM+CHGR6KzTua2lWY01o0Tanj49W1MrzuZqnIvShHjb7nv0nHs3lzP8aeczpGDvbhcDvbu91HU2J5QqPDfn/kHG7f7GVvt5cazT415315CZeXsOsqKXOxs6aSpPcC7nzVx0elHM7TVn+DjqCjx8B+REvP2z27xqbXc/bf66Os/nn8Gw0oLD/GtJ2fdunV09f2Bmbty9dJXiC9ysH7RhLQjshpb/Vy4bH3cdwVPL5gU1RiS7+Ng8bRx1IwYxG0rNyR8Fk+fn1rjSOfe+jP6/g5v0hEcPwLWiYiVLX4sMK8H15wMnC8i3wYKMX0c/4XZmtallAoBVcCuHlwDSJ6ncf+sumiCnx0rXPWGs05kSPHB/INUpT2cDkm6vbK0MOpnsLYtWvsBi6fVArBiVh0Nrf7o5P7SpgabcDp4zA8ee5fFdUJDqx8FHDXYzH+wwoOHlRZQUuCipNBFY5tZgn3j9hbuenELSy4YxQmVxXzS2M4dLxysvnvN6nei9aAsx/qTb3/BzEnHsnrOeAwFHpeDERF/SPxnZ/k4gKjQCocNQiEjJqCgN+mNGkLp1LdKtU9FsYdCt4NVc8bzeVNHNGfmmIoina2uOWxJJwHwBREZCZwU2bRZKZXYLCJNlFI/AX4CENE4blRKzRSRJ4EZmJFVlwPPdPcaFvE5HWBGHE2vq06YjM6qHYY/pLjlj3+nsqSAO2eM5qanPkjpi0gV8hpOUYqkotiDQ6ByUAFDStxUlhSweFotRw4upLQwuXAKG4XMWP5G1OleM6w0ITz4t5fXcf/McVz76LumU7zNz9DSAoo8zoTKtDuafRR5nNG/LYG2u6WTSx7YEL2HpxdMpqLYQ4HLEZNgOMjrYsGUr7DonK/yRVMH//7MP2hs87Nidh1fPWJQVhzFveF4T0f4pNpnWGkBQ7wemtoCMcEO9ta3Gs3hxiEFh4gUAT8GjlFKXSMiI0WkRin1XC+PZRHwBxH5BbAR+F0vnx+FGUq7fN0nCeGqdsd1ZUkBYUPx0JXjKS5wsmzmOBZEJmbL/FPgckSFi327Uoqzaocxva466hhfW7+dIcUerntsI41tfh668nR+8Z2T6QiYdnuryVL8pBUyDnbL+/VfP+bn551MkcfJE3MnUuByEAgrbn32Q8q8Hh684nScDsFQMLTETTBMzDnHVpexcOpIKko8rJhdFw0GqCj20NQeiDr6y7xuAqEwLb5ATOKhNaYHrzg9YqI7uH3e6vqsOYp7o8RIOsInVRfCIwd7U7bR1c5xzeFKOqaqB4F6wEow2AE8CfRYcCil1gHrIn9/Cozv6TntdFU2/ZmNO3n0ajNKylAKp0OoLCkw/Q9xuRL3XTqWVXPG09jqZ1hpAT9e8z73zRxLkccZXZE7xMxXcDlh4dQTY/wV98+q48/v74qajJrbg4QMg86gwY/WfEhlSUHS0iabN5otT8ZWl3H5GcdF/DIHzUY1w0qiGsia+h3RCXFQYWz9qcqSgoQQ4qXTR/Pw69sYUuzhjX/u49bzT+YHjx0Ujqn6Y6Qy0WWzrHlPc0HSET5d7aNLuWs0saQjOE5QSl0iIt8HUEr5xOr5meekKpt+10VjKHQ72NXii5lM7585jkFeNzNtnehMf8NGbv/uKdzyx7+zeFot448tQykYVOimvLgApQxaOkLM+t2bLJ5Wm+CvuDYSxmoxYnABWxvaOaaiCCsD3d5k6ahyL6tf38bJpQU8MXcsQ4o9XPnQ2zHnnP9IPWvmTepyQrS/d0kSv8ujV0/A4xLOHXNkQlmTVP0xClzJizumakGbL6QjfFLto3s1aDSxpPO0B0TESyTKSUROALrt4+hLUq0URwwupLk9GBUa1vZrH30XI0VvikK3k6XTR7O7uZ1pp1ZxyQMbOPPuV7l05QZaO8Pc98rWLp3pJ1QW88TciTx+zQRafCEWP/Mhu1p80Wq7Vt7FDU++j1OEaadWsbvFxyUPbGB/Cmd+KGx02WfZPhEmO97lEJwOB1+2B5K+n6w/RqHbNNHZt8e3qh1oJOufnsrPYvVdCYYN3Q9dM2BJR3D8HHgBqBaRRzHLqt+c1VH1Eqma/hhKUVaUfII3FEmPGT6okIpiN984eQTXJsk4n15XDRB1pscfv/1LUwh0Bo2oGcvpkKSTsEPg2kfqo7kLTe2BpOd0RhpEpcKaxKz944/3uJwEQ0bK81v9MdYvOpOnF0ymZngpnUGDO14wtaMn5k5k8bRa7nhhC77AwDXb2M1Y9s8i3s9imUYvXLaezXvM31v2tmrhoRlwdGmqipikNmNmj0/EzLf4oVJqXx+MrcdUFHtYMauOeTZ/w9Lpo83eGKWFSc0PQmIp9XsuHoNCUVzgJhBKXnnWWn0mc7xbfhUgJlPbIcLtz2+OyTC/44Ut/Pp7p5r7VJHynMtmjmPV69v4zrhqRlaWJNSQgoNlMipLCrjv0rF82R5MCCe1MqTvvmgMNzz5foxjOFmxQ6t/uZWVbn1uA91sk46pK8Y0WpVdJ7qunaXJJV0KDqWUEpH/UUrVAX/uozH1KkOK3dx10RiGlnhwirDnQCfrNu9l9hnHJkyWd180hnZ/iIdf38ZDV46nqc1PRyBMebGHkgJXdFWdTOBURlq/btzewsOvb2PVnPHs9wWpLC3g3r9ujUYs2XNEDKVYOHVkQiFEZyTU1+qQZ53zD3MnEgwrQmGDla99ypr6HWxr6uCH3ziReatj+3rYy2RUlhTQGTSShpNWFHu45Zyv0toZjAm99bqTCwLdgS81feVE760yLBpNd0nHOb5BRE5XSr196F3zi6b2AA+t38a5Y46K9q4wQ0pPw+1wMKTYHTNZFroddAYNrvra8TS2duJ2OjhxeAkjBpstZ7/ET0tHICEMd/msOkTg8Wsm0uILsqvFxw1r3qexzc+T8ydx4bijovufVTuM+2fV8d8vf4xDJKEQotfjRCnF3ReN4b23XgdMwfTDb5yIxyV8bekrMfc4va46KjTg4Cr3ibkTo9vmTzkhKiDt+1gr4ZJCV9LQ22QrZd2BLzWZOtG7qzV0VYNNhwdr+oJ0BMeZwHwR+QxoxzRXKaXU6GwOrDcIhMKMO7YimoMBUFlSwL62ADuaO2OqwoL5kP9h7kSa2wPsawvg9Thx2VpIhgzFvEfejSbvHTPES3GBG0Mp/EGDp975gnPHHBXVHJbNHIdhqBgnvNVG9WfTTubSlbGRTlbEV8iA25/fzJUneHlq/jgqSwvwuh34g4q//vjf2LPfx90vfRzNw0i2ynU4JFrWvaIk+T7WSjiYwvyWaqU8UEql9zZ2bQxaD+lE767WoMODNbkmHcFxTtZHkSU8LidHDCqMecjmTzmBm576IGVZ8z37O6PZ2vdcPAZ75LE1we5o9vHypr3MmnQMV688mFuxbOY4/vz+Tn518Rg+bmjjN3/byk/PrU24zkubGrjlnK+mjPiyIpR2t/g42in4Q+GkocOdQYPhgwoSEg7f/ayJL22Zzg9ecXqXK+HDNdy0t/0Edm3svbfW8/T5k1Oesydaw+H6fcWj/Ty5I2VUlYgUisj/AW4CvoXZcOlz66fPRtgDyr1uyorcMRFDVrhsMGwkjSQa7HUztrqMHc0+Vv7vpwTDBjubO2ho7cQhB6OTrvnX42M0mR3NPhY8+i4zTjs6Wp2xsTWAQyTpdTyRfIj47Zv3tHLJAxtY8p1RHF9ZQkWxhy9ThA63+UM0tga48ewaljy3yTzuuU1cOvHYaEAAwL0vb02I3rKvhDMJNx0o2COgJi99pdcioCxtzO10JIRH2+mJ1nA4fl/xZOv7y+T6ja1+djZ3HJZh112F4z4MnAb8HVPruLtPRtSLNPuC/PLPm7h/Vl30IesIhDmrdhglha6EyfTui8Zw54ubufHsGm74xkh+cOZILnlgA5OXvsJ3l73Olx1BHrry9GgobLIH3+kQtja0seS5Tdz8rRrcTjNKy36dpdNHU+RxsMI2Lmv78nWfREN8O4Pm5GKVdY+/VpHHycI/bGRnc2eMUGls9cfsv3F7C3e8sIXVV43nibkTWXLBKApsCXvphpv2FvmQ65Bqxd/U3uOmk2mRKlQ8Ha2hr7+vfCSX31+uhVY+0JWpqlYpdQqAiPwOeKtvhtQ7GIbCFwzx0qYGfnpubdQJbijFT875KrN//1bUV1HmddMRCGMoZfNB1HLpytgM8nmr63nsmgk8evWE1H29nY7o5H/TUx+w+qrxvLZlb7SWVNhQPPXOF5w4vJTK0gKWXDCKYyqK2NrQxl0vHqxka/opCnlw/TZ+em5t0mtZVX7LitwxrWwtbcq+f2Obn4/3tkXDaOOd3931W2RqLrDb9i+pauW2ZetzEhGUaz9BT6PTDnc/Uy6/Px2c0LXgCFp/KKVC/aTKCHBwctqz3+yH8WV7IKZ0xtMLzoj6KqyJdGx1Gb+6eAxPzJ1IZWlByiq3DQf8tHaGOPmoUu6fVRdNBrRqUr3+z0aA6ETucTq46PSjY6K6ls+qY5DHyZ42P1c+9HZCWXWwKu0aTK+r5pd/3pS0ltVdL26Jmteuf3xjTN7J7y4/jasePjgp3X3RGAyleGLuxGjob08fsu44ePsy16GrcYukEPx95CfQ0Wk9I5d+nlwvOvKBrgTHGBE5EPlbAG/ktRVVNSj1obllX7s/mvi2dPpoWjpiS6M3tPoTKsfe/K0aZkdCUqvKvayaMz7pP+Zgr5sCl4OfPf0ht11wMo9fMxFDKVwOIRA2qK4o5pZzTorJD7lzxuhowUDLDPVY5LhnfjCZsiJ3ghCyihyOKfbw0qYGyrweHp4zHqdDCIUVD7x6MHLr9uc/iln9/GjN+/zXJaey5IJRHDe0mAK3g4YDnfzgsY0xY/J6uveQ2bWMTFdeuX7oLGF3z1+2JAjjvvYTHO5aQ0/IZT6RDk7oQnAopfrtp2D5BnY0+3hm407mTzmB1VeN57N9Hdz78lbe/awpJqN84dSRCc7n25//iPsuHRdTMXb5rDrufHEzPzhzJI2tAZwiON2CLxBm936zq96IwYU0tQW4+6Ix0ZX9TU+ZfS8s7WZHs1ln6jKboPrNpWO566IxDB9UyN4DnSx9fjPfOUIYPqiQs2qHccHYo6LlzK2x/PisE2npCETNaxY7mn0MKy3gmIpiKoo9NPv8fNkeTBjTHxeckfFna9cyUkWmdSUEcv3Q2TWextYAi6fVUlHs4cgyL0dEGlhp8p9camw6CTa9cNx+hWEonCI8NX8SwbCBQyRmgl4xu47SQhe+QIhVc8ajFIgkFgF8aVMD1399JLd/9xSOGFyI02H27G5sDfCDx97lrovG0NwRxB8yotFVZ9UOY+HUE2OS+iyTUpnXHT13VbmXz5s6YgTVdY9tZPG0Wm588n0enjOeX3//VDbVb2DJc/+I6RVi7T//kXpWXzU+5UTs9bioKPawr91Puz8EmLkhjW3+6JiCIYNMsU+8qZpcdSUEMsl1yJR0/C12jccqLAmwftGZvTLpWGOwHP/a/JQ9cqWxaTNjekUO+w3WaviSBzYwY/kbdAaNhIzpeavrCYQUVz1czw1r3qehtdNW4uMgVeVe2gNhRIQrHnybM+96le+vfJP/uOBk7pwxmqOHeHE6HDEhudPrqqMFDK3rLVr7AQunjqTDVq7k/pnjuPflrTHX29F8sLJuW2cIf9Dg86YOXtrUkLI6rlnr6qOEqK0Vs+soK3Tx0Z4DfHfZ65x516ssfuZDbjy7hsqSguiYrAk+k9BC+8Rr1dDKJCzUMBRlXjePXj2Bk44o5ZnrzugVx3i6kS49iWbKZAy6yOHApquq1IcDA0rjsFbDVrTUsUOLUky4RE1Z31/5Jjd8YyTLZ9XFNF+65+IxDBtUGO3NYXXI84cMKksLae0M0RmMzbhOVVL92KHFuBzw1PxJNLUH6Awa0T7hFlXlXoYUezirdhhlRW4aW/2R6riOBJ+Mtb8VBWaZXKyoqqHFHnYf6EwoRWK1ip23up7jhppmrEwc3IahCBsqOpaYHufDSvC6u155hUIGm/e2Rj/nn5xqMKLFz+CCnq/W0o10yaaZIR8c/xpNXzCgNI5AKBzt4Le2fjsqRYl0t9PBWbXDGFtdxiNXjefbY47EFwhz+3dP4an5k1g1ZzxPvrODPfs7o0LDSrKbsfwNrnjwLb5sDyQkF6Yqqb73QCcLH3+PpvZA1GS1YnZiDsedL27m+qkn8tiGz2hqD+Bxml9PspW9VeXXKqw4b3V9NAEwrMwAgGRCrMxrjrmowInDIRnFwze1B/hFJMLLGktjm59hgwqoKvMecuXV0OaP0cgCYbPEfENbz9u7pOt0j8+BWDNvEkOK3DS1B3qsGeTa8a/R9BUDSuPwepzcedEYWjoC3HT2STzx1ucJkTP3XTqOR97Yxq3nn0xTW5BrH42NZPrlnz+isc3P4mm1UUEwf8oJ0XPAwbpSj149Ieb8a+u3J0RHLZ9Vx+L/+TDGng6w4Sdf57FrJhCOrOL37O+ksTUQ7Ra4fN0nXP0VL1Xljmh13EevnsCBzhCFLgedwTBt/lBChd/ls+pwijnJJ+t93hEIs/Ky0xhabK6AM5nsAqFwSg0nHY0hGE5eEysUztzXEk8mTneHQ6go9vR6hdlcO/7zCV0OZGAzYASHYSj2HvDHTARLp4/mmY07Yya5Io+Df60ZzpY9bTFFDuNNOWVeN7c/v5ml00dT4HIknfDANE9Z/ciDYQMHZpVcf8gsTe50CJWlnpgEvbX12zGUYn9HkGsffTdmvHe9uIUjBhWycXsLe4bB4ml1VBR7GDG4ELdLOOBT0TayVeVeVswaF71+iy/IvS9/zK3nj+Ldz5q47usjoz4YK8+kuryQQYXmQxxverIwzWDmZ2p/2K2J0S4ErUTCdHA7k7eddTkzU3yTTUqZmqBSaVpr5k1CKdWtya7c646aPC3H//JZdZTbAiMOB3TZ94HPgDFVJZsIFq39gKm1w1m+7hNafEEqij0UuF0sWvtBTEMlC7sppyMQjq70hw8qTGqC+rSxnQOdQdr9IdbW70BEmPdIPf9yxyvc8cJmQoYiGDb46bm1vPtZU3QMPz23FkMRFRr28S6cOjLq66gsLaCi2ENlaQGrXt/Gx3vaEo6Z98i7fNke4JIHNjBvdT0vbWrAKTBz4rEJtbSufaSeYJjow5vM9FRVbhZr3BMJL7aTbo2kVM72YSUFLLeVWfE4HSyfVcewkvTt/6mc4EBGZThSaVq7WnzdLiPRHBHci6fVcnxlCYun1XLvyx/T7Ase+uABRK7LuWiyjyjVfyM+ioqK1Pjx4wHwhww2ftGcsM/oqjIcAsGwOYkXuB38fcd+ThxeyudNHfhtJpkCl5PjhhbjdIDH5SAUVricDsKGQdhQ7N7fydASs4Cd2ynsavGx3xfiuKHFeD1ONu06gNspVA8pwuNy4A8a7GjuIBhWnDi8hJ0tPr5sD1DgcvLVEaW8FykvYufU6jJEBKUU7a0H2O1zEAwrjq8sptDtYOMXiceMPbqMdn8Yl1MwDCgpcBJWJP08xh5dToHLQTBs3tN721soKXBxZJkXl9NMLixwOQgrRYHLGVPTyiIYNjAUOMTUIuLpCITZsqcVfyhMgctJzRGlFEWSDZUyj1dK0dHWSllZGZkUJQiGDT7ceSDhext11KCkY8n0PMdUFPFxRBBlel77/2B1sWJ7u3lj1mc+kGhpaaGsrCzpe6mexf70OXR1fwOBV199tV4pdVp3j+8f32IaOMR80O0MKfaglOKj3a38Y9d+Pm/qQClz+64WHycMK44eU+BycuLwEhwCnzd1YCjzAVDKLD0iIhxVZuZf/GPXfj7a3UpFSQFup+BwgFIqKjQ+bWzn/e0tbNvXTvWQItxO4eO9bQyNrKz9oTCKxPEWuJyICJt2HeC97S34Q0b0+E8bzVYoye4xFFZ83tTBpl0H2LavnUBYIanOjzmxf7jzAB0Bc2Jv84f4eG8rm3Yd4PMmU9AZhvmZJsPtdFDgciSdUINhIyo0rHvdsqeVYMSPIWIK5QK3ExEyEhpA5HuJ9b/4Q2GMiEDyh4zotbrC7XRQc0RpzPd/fGUxu1oOaiHWedMl2f9ggcuZ8nMcqOjPYeDTr30c1dXVrFu3u9yr1gAAIABJREFUDjBNGB/tORDTQvXRqycw87dvUm4zSRSWe3nj6gk8tuEzvj/hWEKGwimwry2A2yXc9uwmitr8rL5qPA0H/FSWFLD7QCeBkNl61X6u4nIvGyOlQ5wOoTMYZvuXvoRWsI9G/CYr507kkgc2APDb6yZzoNNsUzu9rpqKYg9Dij0YSvGNX70GwPWnhHhiR2n0+Edv+Df2HuiM6cuxas54Lvv9WzHj8pR7eWr+JD5pbE+ob3X80CIuWrGB8mYfJ1SXsTJJeZQij5PSQjfHVhRnbJPe2dzB5LguhQBrFp3JUeVFMdvWrVvHlClTMjp/Y6ufC5etT/CTPHb1BC797Zvstvw5adjU7b4SEeHWZz+keVMDJbbzZhJKG1/A8YkdpQPWtt/VdzcQfBzd+d/sT/S09mC/Fhx2HA6hZlgpj109gYZWf9SemiqK57xTq5j1uzdjJtWV/7uNhVNHUuh2UOh2MKKsEKcIi9Z+wP0zx8U42Zev+4SN21vYe+Bg46f7Z9Xx+Fuf89KmhoSscauarcW+tgAbPmnkB2eOjClrct+l4xhbXRZTJffIwYU8eMXpKAUjBnu55+JTMZSiIxBOmvVuRkUZPPz6tpgxP/z6Nv79vJNjMqdvf34zt3/3FKqHFOEQQUThcToZWtK9pKZsRxYlc4KvmF3HL/68KcGmfqhJ3555bBiKH32zhk27W7ud35FJIyfrmgMx8khnVg98BozgAHC5HFSVF+H1uKgu9xI0VNKQVI/LyeWRarVw0DFtTaCgzFIkCCFDUVlSgKGIVrC1hMLDr2+LEVBWKO1Lmxqi51xywSg6AuFon3EwJ9LyYjffm3AMs38XW0rkB4+9y+3fPYX2QJjQzn/wh7mnUuB2xJQxWTG7jqHFHkTM8SWbqN1OB1dOPi5GO7lzxmg8kRwW+2ey6o3PmF5XzZLnNkVWhkXdfsi7k2CXyQSabFIyDCNpva5M8id6a7KLb+SUioGwKu8KXcBxYDOgBIcdf9jA63Zw/dQTY/Iqls0clzKi5sgyLw0H/JQVu2k4YFaZXTytloVTD2oF1r6L1n7AqjnjuWHN+zHnsNekMrPGi3A5BH/ILJF+1deOp8UX5LZnN/Hv5yW2lbXGcdnv3+KSqjb+c+P7CdV1562u548LzqCpLZCyyqvHJRR5nNE+JB2BMEUeJy4nCZ/J/bPqGDGogDXzJkVzQLq7Qsx0Au7OBBo/KTWmyKzPVMvpy8lO93TQ9GcGjHMcYkM1//WOdbT5w9EJEg62d3WnqFdkRVnt+NIUGjuafSxf9wlHVyQvXdLaGYqalKxz2M1RVeVeWjtDLHluE4VuZ7S967zV9Wzc3hIt955sHPHJhvOnnBBz7c6gwTWr3uGlTQ3c9eIWFk+r5an5k1gzbxI1w0tp84e59dlNkbIlZpb2rc9uot1vJHwm1z5ST0cgzMUr3mDCf/6tW6Go9hBcS/CkU8enN0I3+2MrVZ1lrunPDCjBYZ+ExlaboXTJHs6wobh/5riEEh73vrwVt9MRk+OxcXsLu1t8SSf40kJXQv7D2vrtMa8rSwu4/usj+bI9kNCq9siy2LyGqnIvD15xGm6n8MTciRxTURztfx5fXdeVxLcRNhRKKVp8AbOSb5s/Wopk3up6Gtv8qRtU2UqUZDp596SVZm9MoP2xlWo2iy0eThzuvb9zxYAyVdknoflTTuCzfR3J7f8OoTNocM/FpzK0xENYmSU/Kks9tPiCeOIynJ99bxfLZo6LycK+79JxOCMTfMhQbN7TyiNvfB5jjvI4zUgrr8dJodtJ2FA8eMXp0XIhLR0hfvf/PrX1hCikqT3ILX/8OzuazSKAN559Kg+/vi2muu6dM0bTETR7pze2Brjx7JoYU9WdM0ZzZFlhwpiXzRzHvrbkZp14IVFZUkAgFGZnc8chzU09Mbv0ljO9v9nUdU+HnjPQ/UT5zIASHPZJyF4yxD6pLp9Vh6EU1UMKaWwLxnT9WzZzHH9+fyczTj+a+2eO49pH36WypIALxx3Fb/62NTrBV5YWcN/f/sk5p4xg8TMfcvt3T4k6ztfU72BsdRkLp47E63HhdjrwB0Ns3dtmajIdPsqL3RS6HVz/+HvsaPZFHbsPXnF6TBmUQNhg0Vqzb7nLITy94AwaWv3c8cIWGtv8rJozns+bOpLW0Vo1ZzyGoXjkqglmYqBS/Oblf7K1oS3hM1kxu45f//Xj6OdodUS85IENaT2QPdEaDtcJ9HCOPOqtaDLtJ8odA0pwVBR7op39WnxBGtv8PLNxJw9ecTpOh2AoKC5wcNHyDSyeVhvT59vyfzwxdyIOB7gcwqNXT8DpEP7jT//gpU0N0Qm+qtzLQ1eOx+UUFk+rpbTQxZ0zRnPTUx9QWVLAzd+qiYlmWjZzXEyY7p0zRnPs0OKEyTZVGRSnCAsffy/GnwKw3xfk6CHJ/S8A19n6kC+bOY4WXyBaRuWxyL15XE7Kve6YUNRkHRG7eiB7ojUMtAk0k0ZO/U1L6g16U0vQfqLcMaAEh/Ug3nXRGI6p8PLQlafT2OqPKQp4z8VjqCwpSNk7wyGwry2Y0P+7sTUQk1vR0hE4mL8xcxxPv2sWUzxxWElUi7H2XfDouzFhujc99QGPXzMxYbLtCISTTsAiwvhjzX4g9rBif8hAhKTHxDvYFzz6LmvmTeLn5yUv4GefvMMquR8k1QOZSmso90b6ihxCIAyUCTQ+AfC2Zeu16SSO3tQSdDXi3DGgnOMAToEhxW7a/QZetzNh5fyjNe9zw1knMrS0IKlzsj2QGHW0aO0H/OriMayYXcfY6rIYn8COZh/XPvou55wygnmr67vsg2F/HQgbCcUFq4eYgs1eBPDOGaNp7ghw3qlV0aisJc9tYuHUEzl+aFHSDoDLZ9Ul7TColEoZ6WTvaOZ1u5J+NqkeyGTO6ZGVJWxtbOuWw7y/oov7HZre1BL6YzTdQKHPNQ4RqQZWAUcABvCAUurXIjIEeAI4FvgMuFgplVgp7RAEDUVTW4B1m/dy6cRjU+ZJWBNufEmO1s7kbVobWv0seW4Td84YzZBiD7es/XvM+8cOLeas2mEMLS2IdvqzssuThun6gtEw2jKvO5pncdOTm6Pb9n68kTte2MLdF4+J9hy3OhF2BsMEw4rG1kDMeVp8QYaWeJJ2GHS7HCnty/btbpcjWsokXb9DstyKw83+rE0nh6Y3tYSBZua0k+9VBXJhqgoBNyil3hWRUqBeRP4CXAG8rJS6XURuAW4BFqV7UuuDDhmKB9dv4wdnjuTTxvaUZpyXNjVQ5vXw+DUT2dXioyUykc+fckLSY1oifb+tBk7x+Rtet4OFU0/kcttka2WXXz/1xJis8ftnjqO82M2ic05i6fObaWzzc//MOgIhFQ2hHVtdxpUnwE/P/Soel4Mzjq9ga0Nb0giqO17YEtMf49nrJieYju6cMRp/MJzUvmxpB/Hbn71uMr5A9/5xB/okmuzB1qaTQ9PbwRADxcxppz9Ei+W8rLqIPAP8JvIzRSm1W0RGAOuUUjVdHVtTU6O2bNkS80E/evUENu9pZclzm6JtZGMiiGbV8beP9nLacUMoKXSzr9UfE8lkRRTZndtWzSlLWLx60xRm/vZNKksKWDh1JMcOLUIpuP35j2JKX1hO9EFeF582tlPgcjDY647uZ75/OkVuJ4GwYl+bn7IiN3e8sJnLzziOD995g/98zxF1bh/wHQzVtV9jyQWjuPKht2P+wVp8Ad7fvj+m4OLCqSNj7tU6fs28SVy84o2E7T3RDr5s9ydcv7HNH3PObBSS64uVWqoH2y6AD+cih4ci31fTkNsih6kKefamti4iPSqrnlPBISLHAq8Bo4AvlFJltvealVLlSY6ZC8wFqKysrFuzZg0hQ/FJQxuBsMFJR5QSDCs+aWwDYEiRGT4bMhQhQ9HcHmDYoAKUMr+goSXmP+0XTR0E/v/2zjw8yupe/J8z+2SSkBASUBIBNzTYCEmFiK11u27FUn/E1gq4C4je3vZnXVql9UrtvYh0oa2CtqKyKCr04tVq6UVpb61rjCBEARFagrLFBEgyySzvuX+8C/POvBMSkpBMOJ/nyZOZd95533PmzJzvOd81ruFzuxg+KERMk3hcei4oPXGVIG68//i8oF7tz0jBbr6vOD/IrgO6iqgwx4/bJfC6dFfYuKHbN+0bAFk+D0Ny/dYxn9vFsAI9T9S2vc3k+zR2G98dn9vFiEEhq2hRIiMH5+jtdQurTnk0rvHxLvu5JxZm86nxudjePySHHV+ErTbHjaCqEwYGO1XjIpHWqMY/6pttn43X7cKXUI+hqamJ7Ozsdq6iEzMCG4UQeA4zwSTfd1iBXsekO0n8vpn43C5OKsq2vjPh5iaCoezDtjdT6ejYZSq92T+n3y7AaUNyjvj3mMz555/fJcHRa15VQohsYAXwPSnlgY6m+ZVSPgY8BvqO47zzzmNnQws3z3kdcPGXS8cB8LPf6ZlvF04dwwMv1VLX0Gpdozhfd7U1Dd/6zuFLnFCQhd/jstxvv1VRzJSzh9nLr04eT2GOj5gGVz/2FnUNLkwfg+J8F3OrxqFJbLucRyZX0BbVKMr149l10LJ9LJxaYbgEJ15DsOSmcUx7eC13fElj3oeHhuhvd1fygBFbcagvQRZddBb/8ou/UpwfZOXMSopyAuw92Ma/J61aFl3/JZbXpO44Vl4xnpy9TUkJEcdx6pBsNE10emV4aMVk/2xWzhxPUU7AOu9wqzrnlX1F2hW8833hDzPP7lZ1RuL3LZE37h5npY7v72m5Vf96DqffbnF+kD98o+/YB3vFq0oI4UUXGkullCuNw7sNFRXG/z3p3p9MYvqGeFyS5XOx0EjlkRf0UpjtZ+HUCpZPq2Th1AoKs/Udhzm51+xo5IYn32XOKx+BhJu+ciILp1ZwZfnQ1PKrS99HItKm7hgyIJASkDdz6fs0tcX42ty1zH6plh9cMpIxJXk2l+AxJXksnFrBvKvOxOMWXFxaZLt2cX4Qj0ukeFDNmVRGyO/mz98/l/EnFtDSFkfTpKPHybCCLEcvlEhcS/E+u/OF9YQjGht27qeuIcyGnfvZXt+c1isqMfVDOBpz/GyiscMXWEqks15KR8uuotKFKHqSTPAW6w2vKgH8HvhISvnzhJdeBK4D/tP4v6oj19M0iUSPkN62r5kBQQ8797fym9e28J//T99F3P+NUr5o1r2afG4X93+jFK9b2CaZMSV5XDd+BFc//lbC7qLcykprUpjtp74pkjYjq0sIx8krL8trPb57xXpmTSi14jacbDGPTC6n5u29gLQEhATHGhtmSvRHJpfT3BajvjlCYY7f0eMEsB1zu+DTvc1pJlzNls59blUZeVleBobsq57kncGi689y/GzimkTTZIf12Z0VBEfLOH2sRrsrjg6Z4C3WG6qqc4CpwIdCiA+MYz9CFxjPCSFuAv4JXNWRiyWrMpbePM7aJVx79nDimkZLJJ4yAUrsdSxmnHdSyk7h1qXvW4Znk+9eeAozllRTmO1Pced9dHI5bpdwnLyy/Yc+6rqGMKcUZVPfFGHBlAr2Hmxz3KXMqgiwfNqZloC45asnpRR+emRyOX/5eA+zJpQSiWkMzg2gafrK3snjJHnHEI7EqW+OOLYZ4MkbzsItBLsOtLLojW385IpRELKPQfLOYP6aLSl5suZMKuOnL9fy4JVlHd5ud1YQHK0JPRN+2IrMpq97ix11wSGl/BuQ7hd2YaeuBSmqjMaWqBXvMKwgCxCOaphnp1Wy9OZxaEaCwyyfx3F1O3xQljV5FecHrRTrdQ1hW/zE4NwAzW1R9oejjvEhrdFDq+Ti/CB7DuqZavM8HkoGBh3vHdcCfPuxt6xr/OyPH/HAxFFWzqwBQS8vvPdPLi8bahMmC6dWUJgTSJnInGwGy24ex4rqHalCcEqFzftrblUZM88/uUO5qmp2NCLAtjMyvdJmT9SsaPLYYXYgRyIIBuf6WT6tkriEgNfFoJBzWveuevb09R92T9GZlCqK/kvGpxxJnnCzfG4uLi3i3q+Xsj8cI+B1zv+0a/+hkq9zq8ooyPY5rm7rmyK2CdBMsW7aRqYvrqY4P8iz0yrxe92A4Km/b05RJ31n7DDrmnOrysjP8rKyegflwwus48n39rhg+bRKGsNRVtXsNBInuvG5Xfg9Lu4y6nQkF5mavrja0XXPyWbw05dr+bcLT+VXazZbAmlwboDZL220XItNYTt74hk247aJ085gX1PElgsM4OLSIvY1R6y68D8crTF098G0xu7OrOzTucgOCqVO7j3pJ9+fJ9ZjPaVKJrgRHy0yPuVIspFSCLj9glP4fH8rM5ZUs2u/cy2NxJQhd76wHk1KW7qP4nw9/YeU2GpaPP3mdh5JquXx6ORy4prkoVc/5s7n13Hd+BEp6UEKsn0sn1bJrAmlPPTqJm566j2mnD2CkM/NmtrdKffW1Wnw+799yoK1W7myfCizVm3gop//lVmrNmAqnIbkBtLaAZLrFDjZDFbX7mFgyGuVkq1vjrCvqc2xFGuWz42T+3ZByMfTN45l0fVnsXxaJYuuP4uTi0I8PtVu4Lvv66WW0AA9++/hUnIkpkJpryhUZwzpPZUaJLEuyce7Dva7NCvHckqVrtSc6Y9k9I5DQIoqw+9x89CrH3P3ZadT1xBm3urN/PLbo/ne8g+sc+ZddSYrqutYOLXC2hV43QKvx2Urter1uBiU7bdW0+ZOxtwJROIa2/e18ONVG9nb1GYFCiaqsIpy/PzHKx8xqaLEiu42+awxzD0rP+TRyeUU5fitezeGozz06iYuLWjmx1dUEo1pjokTH77qTPKyvGkN0WYQkbmizvZ7HM+VCNvuYOHUCsfzWowo8mQ0TabYkRZOqSA36LF9nhrOhbW6w+upPUN68kpR0zSbN5uZPNI890hXkbaJtTh9mpVMXbn292wA7aFSuNvJaMEBevDb8mmVfL6/lfrmCI3hKJMqSvhnfYu12vW4hW0Cy/a7uabyBG5fVmOb6B55/ZOUyO9nb6lk9sQzGJTtQwKTjfgQc1cwf80WK6Lc9JaavrjaUmGZWXGnnXuSrd2JaUxuXfo+T94w1maEB7ggTyOuSeLSecIdMiDAgy/XptbXmFLBT1+uTfmS/+Jbox1rcRRl+20CeEX1DhZMqWBGQobguVVlDM4NONoX9jS1Weea95u+pDrFsSCdt1VnvZ46k+4j6HOnqKUWTq1IWwSrK6qXjkysmZBOIh3HckqVY1loOpHxgkMzIsKrFrwJ6KvlgpCPB1/WkxhGYpolIEx0NdRomx3iV2s2M6mixCY46hrC1DdHuOHJdxOC9exGdlNQmMeG5AasNgwM+ViwdivF+UEG5was5Icrqndw3fgRPPynTdb73A7p0X1uF5qU7PjCuZKhxyVYXbuHvQftdpiBIS97D0ZsO6oFa7eiSZmaEDHkw+NxpdgS8oNeVs4cT2tUwy0g6HOTF3ReGUfjmuOPKstnn1Dmr9li1Uupawjjc7s67fXUXroPJ0N6TJMpK8Xpi6tZdvM4tu5tTvFm68oq0ufR7WuTKkqI7dzIwqmjWFG9wzaxZvLKNdFZAQ4eU27Ix7LQdCKjBYcEdu7XB9Ic1DW1u7ntgpOtIk4zzz85ZVIrzPYT8LosQWB6LQ3O9dsm2xXVO9h1QI84T1e/I7kWeF6W1+bhNLeqjJvPHWEZm00X2iVv/sPaqeipSlqt3YCZAytat4HdB9p45cPPU3YKj0wuR6Cv4uev2WJLcrjy1vHcdelIFr2hx3gUhHzM+9aZhCNxy6BvnvuHmecAzl5CToZwJ7xJpXbNa5vlbk32NrUxMKQLpGhM46Oatzu90m5v4nUypH++P+w4bm6X4KSi1GJaXVlF5ge9fPfCU5mxpJpvFzexvKaWBVMqyE/4jmTyyjXRWeGDd97gD984J2PUbF1Fxe7YyWjjeGs0zuTfvU00rlnqh2sqT8DnETxx/Ze5snwo2/Y1pxjHv3vhKdyaFBH+1N+30RbVbEbtf73wVN7fXg9AYzjqaGRPrAW+YEoFD76cuiv5ojlq81CaadTvMN83Z5Ke4fapv2/jmVvGMfubZzBr1QY27T7ID55fx8QxQ1lVoxeKWj6tkmenVfLyup2cM+d1Zq3awF2XjrTqhDx+rZ5+ZtEb2ywjfdWCN7n2iXcArIj07vziF2X7WWBE6id+FiUDgylR7j95cSMCwdD8LDwu0a1V35wM6e1FeXe27sjhaAhHU1R2M5ZU05CQUj/To87Nz9jrdrXrrNDfcKo5kwnqxZ4io3ccfo/b8FL6mAcmjuLHV4yiLaYRiUkGBLzsbGglL8vL0zeOtcUkDCtILbc6qaLEUqGAEQC4pJrFN47l5Q27WbB2q1Ue1lxx/Orq0RyfF+S1O76G3+MiHI07eiMlf7XqGsKcWBji9R+cx/Z9zTz8J72G+F2XjiSucWjyKbZHmpt2k9kTz+DckYN5Z3sjNTsa9biUW8Zh5vuKxjVuOGdESvzK9CXV7VYB7AhO9gWPx8Vpg3N4bvrZxOIaHreLomw/9c1tjrEcP7niyFfX3R0U2J2ryI7sJtTKNXM5VmN3nMhowdEWi7Oiegc/uryUcFT3cJq/Zgtjh+dxTeVwcgIe9hxs49G1n3Db+adw34RSPmtsxSVSo7sLQj7HH/2B1pgV33B8XoAlN42joSXCwJCPtlicqxa8aVMfXVxalGJgT3bYK84P8vn+VoYX6Anx7rnsNMuT6t6vn55WJWau2k1BM7eqjAOtMYpy/MQlPJigDnNKl1LXcKgK4JHQnmHX43FxfJ59Je1yuVJiObq6uu7sxHu4WJDujADviFBTUeeK/kBGCw6XEFw3fgRTfn/I0+k314zBJYTN+2nOpDJ++/oWbvzKiQzJDfCzP6Z6IhXm+B1/9NkBDycHsvlnfQu3La2xJmyvSzD5SbuufebS93n6xrHUfn7QZuM4wVDZJB4L+dxoUjJ/zRbLHdSs7OfUjqIcP7MmlFqr9jEleWT7PSl1Q8za6E7pUro6aXfWsNsTq+sjmXjbWyl25yqyo8ZjtXJVZDoZLTg8bpHiFdPQHLUVK0pU9RTl+DnYFnP0RIoaNcCTc08dCEf412c+sE3kZhVAp53B/nDUdt2HXt3E/O+MTonR2NvUxsoZZ/PLq0ez92Cb5W11x8Wn8uiUCm5dUo05+Tw6pYJn3/kHC/93u5VF97QhOZZwTO6nGWSXnC7lSDyYEtVSnTXs9tTqujsm3p6IpTiWjceKY4uMFhyC1Ey0eVnO3k8FIR9+j15IqDg/mOJdNGtCKWtqd7Po+rP0AkxuF0ve3MZVZw1L65XjtDNobInaAv2K84O4hUiJ0RhTkseepogtVmLOpDIOhOM8/r9bmTWhlNjOjcwaXcqv12zmnstOZ1t9C9eNH8FTf99mBTgmt8v08irOD7KzIWyraX7cAH+HJ8tYTOOz/XqtdVOo3ff10k67JPbF1XVPxlIkG48Viv5IRntVmSnHE8kJeB29Vgpz/Ny+rIYH/nsjjyZ5AM2ZVMaa2t1MHDOUG558lwvm/YXvPP4W544cTMjv7AXzeWOrdZ0xJXksuv4sFt80lhGDQjbPpblVZbgMIZOImWU3eccwKNvH6to9TF9czad7m5i+uJrVtXtojca5b8Io/B4Xd15ympXWPbldpvfXnEllzFu92UqXMn/NFnY2tHYoZYKmSTbtOcg1v3ubqgVvMvulWq4bP4Klb223osrN+2WiYfdYTp2hUHQHGb3jcAlhqXXqGsJM/+pwsv1uHp1cbrnbmq6hfo++kjQN10/fOBaXSy9F+sB/b2RSRUmK2uvuFXoW3WRvKjNTbWGOj6U3j+NAOGrd7+LSIn50eSk/vPx0YnHJ43/9lNsvODnlGic4eHbVNYTRpHPCw5DfwzUJtUJ+c80Y5l11Jnc8v87Wz/wsL89NP5v7X9xgxYmALqiSvcbS2SfqExIRJn4WsyaUMijk67Dqqa+m1sjkWAqFoi+Q0YJDAr9es5nFN44lpmm0RvUI8sJsPe/T8EFZ7D7Qxqz/2mDLJbW6dg83feVEhgwIUN/UxncvPJXWqPNkEtckHpeLZ26p5LPGsM2tFOCHl51uCQ2zGFSisX5ulV6hb3BugCdvGEtjS4T65ogty66JrvZxpRhYF05NjQ+5fVkNc6vKbKqog616HMGLt5/D9/9lpM1IP2JQx4Pd0k2s+sTfMRVMX06toaKAFYqukeGqKn0HsXlPE0K4uHWpvko2S8FO/f077A9HqdnRaK2aZ5x3khW4F9ckty2rIeB1MTDkc1T9eAyddSyuccfz65i+uNoW8R2Xh0rIOhWDuvOF9cQ1GF4QIj/kIeB1M/ulWuat3szcKnsZ2IVTKzguN2AZWE8bov8fFPI5xod43S6mL67mjufX4fO4eOjVTdQ1hAlH4inBSllpVG5Ok2W6ILWiHH+H1VJ9WR2UCaU5FYq+TEbvOMCIUl67lV9cPbpdY7H5vCDkY25VGYOy9TxSdQ1hrl/0LstuGeeY2M/rdrF59wGeeecfKV5XC6ZUsK/pUPW8dGlJWmMaLpegIBQgP+i3VD1et4vnplUScyg6lGhgTVem9rgBAVbddg67DrRauyBTGCQbpTVNdtg11smNduHUCo4fEOyxsq9HExVLoVB0jYwWHAIsG0dck46Ta2NCuofi/CBDBgSQUvLrNZ/wXHUdoE9osbjE48aWRbcwx4/bDQNDXm44ZwSL3thmBQMW5vgJ+lzkxTyWwDEN08ltcCfMR+15GWmatKrjJe4E0sVDDM4JcKA1Zsu5lU4YdGay7I6Jta+rg/qit5dCkSlktOCQQLZftz9k+Vw2Q7kZ//DSB7pwKM7X0437PIKNO5uYVFHMhaWDWbB2K3ub2gCIa3ByUTaN4SgtkThSSlojGgOCPnKDkh9PKKUtJvF5XAzJ8eN2u6hvjjAk183yaZUQdwXJAAALr0lEQVQIgeOuJeg7/GTpZBO4rxyrPkS6ibynVs5dnVhVag2Fov+S0YID4D/++DGTKkoYdVwuv15jL9n66zWbufOS07jg9CG0ROIMzvWzvb7FVnBoblUZA0M+wtE4fo+b7z5TY0Vm33XpSO584T2baqoox8/ALD0/E5CiDmqNarZdy+DcAHnBw0+WTjaBf9Rr1DdHrGRyThN5Ryf4o22sVuoghaL/kvHGcTMDbFTTrPgHs8zr6to9fNEc4Y7n1zEox09rTLMqAcIh43VDSxQp4c7n11mG7xnnnZSSJHDGkmqEEJbQSMblEgwvCHHG0AEU5wc5Y+gAhheEOjRZOtkEInGt22wCvWGs7mjZV4VCkVlk/I7DNFanqwkxODfArAl69PVdlzpHWxfm+PmiKWKprACOHxCw7V4WrN1KzY7Gw07kR6ricbIJ+NyubrMJ9GVjtUKhyCwyfsdhToZel+CRyeU2F8tHJpezPxyxdh9mlb1EivOD/LO+hfyQl19860wrElyCrTbHDy4ZycWlRT1m3HVyER1WEOo2m0Cm14FQKBR9h4zfcZilOqNxyW9e22LbJfzmtS388LLTAX2S3NcUsZUuNQ3oeUEPbTENn8elBw4WZDH1iXdSIqeX3Tyux4y7TjaBjdVvdpt6RxmrFQpFd5HRgkOAVapz2S3jmFRRwvEDAgS8bnweF5MqSnC7hWUEH5DlITvg4eGrzkQALZE4Qa8Lv8eF3+PGheCL5igSHNU6AHWNLQS8blvMRXfRky6iylitUCi6i4wWHBK9Wl5htp/GFr1G+HXjR9jyVC2cUsELM84GdEEza9WGlEJLsyaUMvulWp6+cSyjhuYSjWmO9pKte5u54cl3+1T6jM6gYhcUCkV30C9sHDPOO4mZS993TFQ4fUk12+tbiMQ1djSEHVN3mBHf1z7xDgLBcQOCKfaGuVVlzF+zxXpPX0mfoVAoFEebjN5xgD6pmxN/upQfRTl+GpojaavrmdHlppdRolonHIkR1yT//7l1tmyzyiNJoVAcq2T8jmNuVRktkbitFkUixflBPmsMc9uyGtwuwZxJ9sSCcyaVsWDtVuu56WVkqnWCPg/b61tsrrrJ5yoUCsWxREYLDoEeuZ3t9/Do5HJWVO9wFAzzVm+mriGMSwge/tMmZk0o5YUZZ/PstEqe+vs2Kzmgk5dRQcjHsIKslEy2yiNJoVAcq2S8qmr4wBBZvjZAMmvCKFwCnrmlki+aIylZY1sicWp2NDL7pVoev/bLDMkJ8OCVZfzkivReRmY0eF6Wl+XTKok7ZLJVKBSKY4mMFxwej4vj84JWtTlN0/S0IC5hyxq7cGoFQwb4eePu821CoiNeRi6XYGDID6Gj0CGFQqHo42S84DBJFgKDsv3OMQtq8lcoFIou0W8ERzIqZkGhUCh6how2jisUCoXi6NOnBIcQ4lIhxCYhxCdCiHt6uz0KhUKhSKXPCA4hhBv4LXAZUAp8RwhR2rutUigUCkUyfUZwAGOBT6SUn0opI8CzwMRebpNCoVAokuhLxvGhwI6E53XAuOSThBDTgGnG0zYhxIaj0LbeYhCwr7cb0YP05/71576B6l+mM7Irb+5LgsMpmk6mHJDyMeAxACHEe1LKL/d0w3oL1b/MpT/3DVT/Mh0hxHtdeX9fUlXVASUJz4uBz3qpLQqFQqFIQ18SHO8CpwghRgghfMDVwIu93CaFQqFQJNFnVFVSypgQ4nbgT4AbeEJKufEwb3us51vWq6j+ZS79uW+g+pfpdKl/QsoUM4JCoVAoFGnpS6oqhUKhUGQASnAoFAqFolNkrODob+lJhBDbhRAfCiE+MF3lhBADhRB/FkJsMf7n93Y7O4oQ4gkhxJ7EOJt0/RE6842xXC+EKO+9lneMNP27Xwix0xjDD4QQlye89kOjf5uEEJf0Tqs7jhCiRAjxuhDiIyHERiHEvxnHM34M2+lbvxg/IURACPGOEGKd0b9/N46PEEK8bYzdcsMJCSGE33j+ifH68MPeREqZcX/oxvOtwImAD1gHlPZ2u7rYp+3AoKRjDwH3GI/vAeb0djs70Z9zgXJgw+H6A1wOvIIey1MJvN3b7T/C/t0P/MDh3FLjO+oHRhjfXXdv9+Ew/TsOKDce5wCbjX5k/Bi207d+MX7GGGQbj73A28aYPAdcbRxfANxqPJ4JLDAeXw0sP9w9MnXHcaykJ5kIPGU8fgr4Zi+2pVNIKf8KfJF0OF1/JgJPS523gDwhxHFHp6VHRpr+pWMi8KyUsk1KuQ34BP073GeRUn4upXzfeHwQ+Ag9u0PGj2E7fUtHRo2fMQZNxlOv8SeBC4AXjOPJY2eO6QvAhUKIdsubZqrgcEpP0t7AZwISWC2EqDbSqgAMllJ+DvqXHSjqtdZ1D+n605/G83ZDVfNEgmoxo/tnqC7GoK9c+9UYJvUN+sn4CSHcQogPgD3An9F3SY1SyphxSmIfrP4Zr+8HCtq7fqYKjg6lJ8kwzpFSlqNnB75NCHFubzfoKNJfxvNR4CRgNPA5MM84nrH9E0JkAyuA70kpD7R3qsOxPt1Hh771m/GTUsallKPRM3CMBU53Os343+n+Zarg6HfpSaSUnxn/9wB/QB/s3eZ23/i/p/da2C2k60+/GE8p5W7jB6sBj3NInZGR/RNCeNEn1qVSypXG4X4xhk5962/jByClbATWots48oQQZtB3Yh+s/hmvD+AwathMFRz9Kj2JECIkhMgxHwMXAxvQ+3Sdcdp1wKreaWG3ka4/LwLXGp45lcB+Ux2SSSTp9K9EH0PQ+3e14b0yAjgFeOdot68zGDru3wMfSSl/nvBSxo9hur71l/ETQhQKIfKMx0HgInQ7zutAlXFa8tiZY1oFvCYNS3laetsDoAueA5eje0NsBe7t7fZ0sS8nonttrAM2mv1B1zOuAbYY/wf2dls70adn0Lf7UfQVzU3p+oO+Vf6tMZYfAl/u7fYfYf8WG+1fb/wYj0s4/16jf5uAy3q7/R3o31fQ1RXrgQ+Mv8v7wxi207d+MX5AGVBj9GMD8GPj+InoAu8T4HnAbxwPGM8/MV4/8XD3UClHFAqFQtEpMlVVpVAoFIpeQgkOhUKhUHQKJTgUCoVC0SmU4FAoFApFp1CCQ6FQKBSdQgkOxTGDEKIgIfPprqRMqL5uvM9FQoj9xnU/EkLcm+a8EiHE8u66r0JxtFDuuIpjEiHE/UCTlPLhpOMC/XehdeHaFwG3Sym/aaS1WA9cKaVcl3CORx7KG6RQZBRqx6E45hFCnCyE2CCEWAC8D5QIIRoTXr9aCPE74/FgIcRKIcR7Rs2DyvauLfUspe8DJwkhbhZCPCuEeAl4xbjvB8Z1PUKIXxjtWC+EmGkcP0sI8Rcj+eUrQojBxvHvCyFqjZoLS3rkg1Eo0uA5/CkKxTFBKXCDlHJGQj4fJ+YDD0kp3zIyq74EnJHuZCFEIXrOo3uBrwJnA6OllA1CiJMTTr0VOB44U0oZF3rBJD/wK+AbUsp9QojJwGxgGnAXMExKGTHTSygURwslOBQKna1Sync7cN5FwMiEcgX5QoiglDKcdN75QogaQANmSyk3CSG+CqyWUjakue4vpZRxACnlF0KI0cAo4H+M+7nR05uAnppmiRBiFfBfHe+mQtF1lOBQKHSaEx5r2FNNBxIeC2Cs1AuItcfrUkqnwlvNDsfM6yYbHAWwXkr5VYfzLwG+hl6E5z4hxBmm0FEoehpl41AokjAM4w1CiFOEEC70TKkm/wPcZj4xdgXdwWrgViGE27juQKAWGCqEGGsc8wkhRhnnFEspXwPuBAqBrG5qh0JxWJTgUCicuRt4FT0DbF3C8duAcwwDdi1wSzfdbyGwC1gvhFgHfEtK2Yae5vrnxrEaYBy6pmCZEGI9uuF9jtRLoCoURwXljqtQKBSKTqF2HAqFQqHoFEpwKBQKhaJTKMGhUCgUik6hBIdCoVAoOoUSHAqFQqHoFEpwKBQKhaJTKMGhUCgUik7xfyZvsvCjTW2eAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.338\n",
      "RMSLE: 35.544\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.386\n",
      "RMSLE: 34.264\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.392\n",
      "RMSLE: 33.525\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.084\n",
      "RMSLE: 49.449\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.302\n",
      "RMSLE: 35.416\n",
      "Prediction of default model: 26.84 $\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "R_2 score: 0.389\n",
      "RMSLE: 34.554\n",
      "Prediction of default model: 26.84 $\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>name</th>\n",
       "      <th>unprocessed_text</th>\n",
       "      <th>KNNRegressor</th>\n",
       "      <th>RandomForestRegressor</th>\n",
       "      <th>Ridge</th>\n",
       "      <th>Lasso</th>\n",
       "      <th>Linear Regression</th>\n",
       "      <th>XG Boost</th>\n",
       "      <th>Light GBM</th>\n",
       "      <th>Actual Price</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6252</th>\n",
       "      <td>N64 super mario 64 (label stain)</td>\n",
       "      <td>n64 super mario 64 (label stain) great item gr...</td>\n",
       "      <td>17.72</td>\n",
       "      <td>14.87</td>\n",
       "      <td>9.92</td>\n",
       "      <td>10.68</td>\n",
       "      <td>14.23</td>\n",
       "      <td>15.98</td>\n",
       "      <td>13.40</td>\n",
       "      <td>21.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4684</th>\n",
       "      <td>Jordan 9c Retro Gamma 11s</td>\n",
       "      <td>jordan 9c retro gamma 11s size 9c gamma gamma ...</td>\n",
       "      <td>19.74</td>\n",
       "      <td>24.65</td>\n",
       "      <td>24.07</td>\n",
       "      <td>21.92</td>\n",
       "      <td>35.60</td>\n",
       "      <td>17.86</td>\n",
       "      <td>21.67</td>\n",
       "      <td>29.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1731</th>\n",
       "      <td>5 Pair of Lashes + Lash Tweezers</td>\n",
       "      <td>5 pair of lashes + lash tweezers perfect for t...</td>\n",
       "      <td>11.24</td>\n",
       "      <td>11.39</td>\n",
       "      <td>5.30</td>\n",
       "      <td>7.59</td>\n",
       "      <td>5.94</td>\n",
       "      <td>13.08</td>\n",
       "      <td>11.67</td>\n",
       "      <td>7.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4742</th>\n",
       "      <td>Bundle for LizKay</td>\n",
       "      <td>bundle for lizkay gb girls size 10. fits like ...</td>\n",
       "      <td>15.44</td>\n",
       "      <td>19.64</td>\n",
       "      <td>22.27</td>\n",
       "      <td>21.16</td>\n",
       "      <td>25.10</td>\n",
       "      <td>19.21</td>\n",
       "      <td>19.22</td>\n",
       "      <td>40.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4521</th>\n",
       "      <td>New Balance Minimus size 8 - women</td>\n",
       "      <td>new balance minimus size 8 - women good used c...</td>\n",
       "      <td>24.32</td>\n",
       "      <td>26.33</td>\n",
       "      <td>34.00</td>\n",
       "      <td>31.79</td>\n",
       "      <td>23.03</td>\n",
       "      <td>27.76</td>\n",
       "      <td>24.85</td>\n",
       "      <td>16.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6340</th>\n",
       "      <td>PINK Fanny pack</td>\n",
       "      <td>pink fanny pack never been used, pink fanny pa...</td>\n",
       "      <td>22.48</td>\n",
       "      <td>11.78</td>\n",
       "      <td>15.51</td>\n",
       "      <td>16.71</td>\n",
       "      <td>8.88</td>\n",
       "      <td>14.97</td>\n",
       "      <td>15.92</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>576</th>\n",
       "      <td>OFRA BEVERLY HILLS HIGHLIGHTER</td>\n",
       "      <td>ofra beverly hills highlighter swatched full size</td>\n",
       "      <td>13.84</td>\n",
       "      <td>11.53</td>\n",
       "      <td>13.74</td>\n",
       "      <td>14.04</td>\n",
       "      <td>16.09</td>\n",
       "      <td>13.75</td>\n",
       "      <td>14.05</td>\n",
       "      <td>19.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5202</th>\n",
       "      <td>Victoria secret swim small</td>\n",
       "      <td>victoria secret swim small victoria secret swi...</td>\n",
       "      <td>15.83</td>\n",
       "      <td>15.15</td>\n",
       "      <td>16.76</td>\n",
       "      <td>16.84</td>\n",
       "      <td>21.17</td>\n",
       "      <td>17.03</td>\n",
       "      <td>15.21</td>\n",
       "      <td>10.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6363</th>\n",
       "      <td>MAD MATTE w/liner</td>\n",
       "      <td>mad matte w/liner liquid matte sold as a set o...</td>\n",
       "      <td>15.97</td>\n",
       "      <td>13.90</td>\n",
       "      <td>10.24</td>\n",
       "      <td>12.45</td>\n",
       "      <td>6.51</td>\n",
       "      <td>15.06</td>\n",
       "      <td>13.62</td>\n",
       "      <td>6.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>439</th>\n",
       "      <td>Jaclyn Hill Palette</td>\n",
       "      <td>jaclyn hill palette brand new! authentic!! abs...</td>\n",
       "      <td>24.94</td>\n",
       "      <td>32.15</td>\n",
       "      <td>32.95</td>\n",
       "      <td>32.13</td>\n",
       "      <td>36.63</td>\n",
       "      <td>26.62</td>\n",
       "      <td>36.47</td>\n",
       "      <td>69.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                        name  \\\n",
       "train_id                                       \n",
       "6252        N64 super mario 64 (label stain)   \n",
       "4684               Jordan 9c Retro Gamma 11s   \n",
       "1731        5 Pair of Lashes + Lash Tweezers   \n",
       "4742                       Bundle for LizKay   \n",
       "4521      New Balance Minimus size 8 - women   \n",
       "6340                         PINK Fanny pack   \n",
       "576           OFRA BEVERLY HILLS HIGHLIGHTER   \n",
       "5202              Victoria secret swim small   \n",
       "6363                       MAD MATTE w/liner   \n",
       "439                      Jaclyn Hill Palette   \n",
       "\n",
       "                                           unprocessed_text  KNNRegressor  \\\n",
       "train_id                                                                    \n",
       "6252      n64 super mario 64 (label stain) great item gr...         17.72   \n",
       "4684      jordan 9c retro gamma 11s size 9c gamma gamma ...         19.74   \n",
       "1731      5 pair of lashes + lash tweezers perfect for t...         11.24   \n",
       "4742      bundle for lizkay gb girls size 10. fits like ...         15.44   \n",
       "4521      new balance minimus size 8 - women good used c...         24.32   \n",
       "6340      pink fanny pack never been used, pink fanny pa...         22.48   \n",
       "576       ofra beverly hills highlighter swatched full size         13.84   \n",
       "5202      victoria secret swim small victoria secret swi...         15.83   \n",
       "6363      mad matte w/liner liquid matte sold as a set o...         15.97   \n",
       "439       jaclyn hill palette brand new! authentic!! abs...         24.94   \n",
       "\n",
       "          RandomForestRegressor  Ridge  Lasso  Linear Regression  XG Boost  \\\n",
       "train_id                                                                     \n",
       "6252                      14.87   9.92  10.68              14.23     15.98   \n",
       "4684                      24.65  24.07  21.92              35.60     17.86   \n",
       "1731                      11.39   5.30   7.59               5.94     13.08   \n",
       "4742                      19.64  22.27  21.16              25.10     19.21   \n",
       "4521                      26.33  34.00  31.79              23.03     27.76   \n",
       "6340                      11.78  15.51  16.71               8.88     14.97   \n",
       "576                       11.53  13.74  14.04              16.09     13.75   \n",
       "5202                      15.15  16.76  16.84              21.17     17.03   \n",
       "6363                      13.90  10.24  12.45               6.51     15.06   \n",
       "439                       32.15  32.95  32.13              36.63     26.62   \n",
       "\n",
       "          Light GBM  Actual Price  \n",
       "train_id                           \n",
       "6252          13.40          21.0  \n",
       "4684          21.67          29.0  \n",
       "1731          11.67           7.0  \n",
       "4742          19.22          40.0  \n",
       "4521          24.85          16.0  \n",
       "6340          15.92          10.0  \n",
       "576           14.05          19.0  \n",
       "5202          15.21          10.0  \n",
       "6363          13.62           6.0  \n",
       "439           36.47          69.0  "
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from sklearn.metrics import r2_score\n",
    "from sklearn.linear_model import Ridge,Lasso, LinearRegression\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "import xgboost as xgb\n",
    "import lightgbm as lgb\n",
    "from sklearn.neighbors import KNeighborsRegressor\n",
    "knn = KNeighborsRegressor(n_neighbors=10, algorithm=\"kd_tree\", weights=\"uniform\")\n",
    "ridge = Ridge(alpha=20)\n",
    "lasso = Lasso(alpha =0.0005)\n",
    "lr = LinearRegression()\n",
    "rf=RandomForestRegressor(max_depth=100, n_estimators=1000)\n",
    "xgBoost = xgb.XGBRegressor(random_state=42, silent=True)\n",
    "lgbm = lgb.LGBMRegressor()\n",
    "\n",
    "#############################\n",
    "\n",
    "result=df[[\"name\",\"unprocessed_text\",\"price\"]].iloc[idx_test]\n",
    "display_model_predictions(knn, X_train, y_train, X_test, y_test, idx_train, idx_test, \"KNNRegressor\")\n",
    "display_model_predictions(rf, X_train, y_train, X_test, y_test, idx_train, idx_test, \"RandomForestRegressor\")\n",
    "display_model_predictions(ridge, X_train, y_train, X_test, y_test, idx_train, idx_test, \"Ridge\")\n",
    "display_model_predictions(lasso, X_train, y_train, X_test, y_test, idx_train, idx_test, \"Lasso\")\n",
    "display_model_predictions(lr, X_train, y_train, X_test, y_test, idx_train, idx_test, \"Linear Regression\")\n",
    "display_model_predictions(xgBoost, X_train, y_train, X_test, y_test, idx_train, idx_test, \"XG Boost\")\n",
    "display_model_predictions(lgbm, X_train, y_train, X_test, y_test, idx_train, idx_test, \"Light GBM\")\n",
    "result[\"XG Boost\"] = result[\"XG Boost\"].apply(lambda x: round(float(str(x)),2))\n",
    "price = result[\"price\"]\n",
    "result = result.drop([\"price\"], axis=1)\n",
    "result[\"Actual Price\"] = price\n",
    "display(result.head(10))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "coeffs = pd.DataFrame({'Feature_name':strat1columns,\"Ridge\": ridge.coef_}).sort_values(by=\"Ridge\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "# coeffs = pd.DataFrame({'Feature_name':strat1columns,'Ridge_Coeff':ridge.coef_,'Lasso_Coeff':lasso.coef_, 'LR_Coeff':lr.coef_, \"xgBoost Feature Imp\":xgBoost.feature_importances_, \"lgbm f Imp\":lgbm.feature_importances_}).sort_values(by=\"Ridge_Coeff\", ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "coeffs[[\"Ridge\"]] = coeffs[[\"Ridge\"]].apply(lambda x: round(x,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "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>Feature_name</th>\n",
       "      <th>Ridge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>Brand_Apple</td>\n",
       "      <td>0.288</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Feature_name  Ridge\n",
       "56  Brand_Apple  0.288"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coeffs[coeffs[\"Feature_name\"].str.contains(\"Brand_Apple\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "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>item_condition_id</th>\n",
       "      <th>shipping</th>\n",
       "      <th>Brand_47 Brand</th>\n",
       "      <th>Brand_7 For All Mankind®</th>\n",
       "      <th>Brand_90 Degree By Reflex</th>\n",
       "      <th>Brand_A Bathing Ape</th>\n",
       "      <th>Brand_A Plus Child Supply</th>\n",
       "      <th>Brand_A+D</th>\n",
       "      <th>Brand_AG Adriano Goldschmied</th>\n",
       "      <th>Brand_ALDO</th>\n",
       "      <th>...</th>\n",
       "      <th>years</th>\n",
       "      <th>yellow</th>\n",
       "      <th>yes</th>\n",
       "      <th>yeti</th>\n",
       "      <th>yoga</th>\n",
       "      <th>york</th>\n",
       "      <th>youth</th>\n",
       "      <th>zara</th>\n",
       "      <th>zip</th>\n",
       "      <th>zipper</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>0 rows × 3118 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [item_condition_id, shipping, Brand_47 Brand, Brand_7 For All Mankind®, Brand_90 Degree By Reflex, Brand_A Bathing Ape, Brand_A Plus Child Supply, Brand_A+D, Brand_AG Adriano Goldschmied, Brand_ALDO, Brand_ALEX AND ANI, Brand_ALO Yoga, Brand_AND1, Brand_API, Brand_ART, Brand_ASICS, Brand_ASOS, Brand_ASUS, Brand_Abbott, Brand_Abercrombie & Fitch, Brand_Acacia Swimwear, Brand_Adagio, Brand_Adidas, Brand_Adrianna Papell, Brand_Aeropostale, Brand_Affliction, Brand_Air Force, Brand_Air Jordan, Brand_Air Wick, Brand_Alfani, Brand_All, Brand_All American, Brand_Almost Famous, Brand_Alo, Brand_Alpinestars, Brand_Altec Lansing, Brand_Always, Brand_AmazonBasics, Brand_Ambiance Apparel, Brand_Ameda, Brand_American Apparel, Brand_American Boy & Girl, Brand_American Eagle, Brand_American Fighter, Brand_American Girl ®, Brand_American Rag, Brand_Anastasia Beverly Hills, Brand_Angels, Brand_Angie, Brand_Angry Birds, Brand_Ann Taylor, Brand_Ann Taylor LOFT, Brand_Anne Geddes, Brand_Anne Klein, Brand_Anthropologie, Brand_Antonio Melani, Brand_Apple, Brand_Apt., Brand_Apt. 9, Brand_Aqueon, Brand_Ardell, Brand_Arden B, Brand_Ariat, Brand_Arizona, Brand_Arizona Jean Company, Brand_As Seen on TV, Brand_Athleta, Brand_Aveda, Brand_Aveeno, Brand_Avent, Brand_Avenue, Brand_Avia, Brand_Avon, Brand_Aéropostale, Brand_BB, Brand_BCBG, Brand_BCBGMAXAZRIA, Brand_BCBGeneration, Brand_BISOU BISOU, Brand_Babolat, Brand_Baby Einstein, Brand_Baby Jogger, Brand_Baby Phat, Brand_Bailey 44, Brand_Banana Republic, Brand_Bandai, Brand_Barbie, Brand_Bare Escentuals, Brand_Basic Editions, Brand_Bass, Brand_Bath & Body Works, Brand_Bauer, Brand_Beats, Brand_Beats by Dr. Dre, Brand_Bebe, Brand_Belkin, Brand_Ben Nye, Brand_Benefit, Brand_Berkley, Brand_Betsey Johnson, ...]\n",
       "Index: []\n",
       "\n",
       "[0 rows x 3118 columns]"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(columns=strat1columns)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "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>Feature_name</th>\n",
       "      <th>Ridge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>976</th>\n",
       "      <td>Sub_Cat_2_Cell Phones &amp; Smartphones</td>\n",
       "      <td>0.535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>134</th>\n",
       "      <td>Brand_Burberry</td>\n",
       "      <td>0.369</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>851</th>\n",
       "      <td>Sub_Cat_1_Shoes</td>\n",
       "      <td>0.349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>409</th>\n",
       "      <td>Brand_Louis Vuitton</td>\n",
       "      <td>0.336</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1479</th>\n",
       "      <td>authentic</td>\n",
       "      <td>0.312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987</th>\n",
       "      <td>gucci</td>\n",
       "      <td>0.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>606</th>\n",
       "      <td>Brand_SeneGence</td>\n",
       "      <td>0.251</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>895</th>\n",
       "      <td>Sub_Cat_2_Athletic</td>\n",
       "      <td>0.237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1906</th>\n",
       "      <td>flawless</td>\n",
       "      <td>0.237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2534</th>\n",
       "      <td>priority</td>\n",
       "      <td>0.234</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             Feature_name  Ridge\n",
       "976   Sub_Cat_2_Cell Phones & Smartphones  0.535\n",
       "134                        Brand_Burberry  0.369\n",
       "851                       Sub_Cat_1_Shoes  0.349\n",
       "409                   Brand_Louis Vuitton  0.336\n",
       "1479                            authentic  0.312\n",
       "1987                                gucci  0.275\n",
       "606                       Brand_SeneGence  0.251\n",
       "895                    Sub_Cat_2_Athletic  0.237\n",
       "1906                             flawless  0.237\n",
       "2534                             priority  0.234"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coeffs.head(20).sample(10).sort_values(by=[\"Ridge\"],ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pprint"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "('[{\"name\":\"Xbox 360 bundle 2 white '\n",
      " 'controllers\",\"item_condition_id\":0.5,\"brand_name\":\"Xbox\",\"price\":13.0,\"shipping\":0,\"item_description\":\"Bundle '\n",
      " 'for SunCity. 2 Xbox controllers. This is a white Microsoft Xbox 360 '\n",
      " \"controller. Works perfect but the battery door in back missing. I don't have \"\n",
      " 'the power cord that plugs in to use but Tested. Comes from smoke and pet '\n",
      " 'free home.\",\"general_cat\":\"Electronics\",\"subcat_1\":\"Video Games & '\n",
      " 'Consoles\",\"subcat_2\":\"Accessories\",\"unprocessed_text\":\"xbox 360 bundle 2 '\n",
      " 'white controllers bundle for suncity. 2 xbox controllers. this is a white '\n",
      " 'microsoft xbox 360 controller. works perfect but the battery door in back '\n",
      " \"missing. i don't have the power cord that plugs in to use but tested. comes \"\n",
      " 'from smoke and pet free home.\",\"processed_text\":\"xbox 360 bundle 2 white '\n",
      " 'controllers bundle for suncity  2 xbox controllers  this is a white '\n",
      " 'microsoft xbox 360 controller  works perfect but the battery door in back '\n",
      " 'missing  i don t have the power cord that plugs in to use but tested  comes '\n",
      " 'from smoke and pet free home \",\"lemmatized_text\":\"xbox 360 bundle 2 white '\n",
      " 'controller bundle for suncity 2 xbox controller this is a white microsoft '\n",
      " 'xbox 360 controller work perfect but the battery door in back missing i don '\n",
      " 't have the power cord that plug in to use but tested come from smoke and pet '\n",
      " 'free home\"}]')\n"
     ]
    }
   ],
   "source": [
    "dummy = df.sample(1)\n",
    "pprint.pprint(dummy.to_json(orient='records'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "976     False\n",
       "2228    False\n",
       "2223    False\n",
       "363      True\n",
       "134      True\n",
       "        ...  \n",
       "1148    False\n",
       "1007    False\n",
       "870     False\n",
       "0       False\n",
       "2857    False\n",
       "Name: Feature_name, Length: 3118, dtype: bool"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coeffs[\"Feature_name\"].str.contains(\"Brand\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "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>Feature_name</th>\n",
       "      <th>Ridge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>976</th>\n",
       "      <td>Sub_Cat_2_Cell Phones &amp; Smartphones</td>\n",
       "      <td>0.535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2228</th>\n",
       "      <td>lululemon</td>\n",
       "      <td>0.442</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2223</th>\n",
       "      <td>lularoe</td>\n",
       "      <td>0.393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>851</th>\n",
       "      <td>Sub_Cat_1_Shoes</td>\n",
       "      <td>0.349</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1479</th>\n",
       "      <td>authentic</td>\n",
       "      <td>0.312</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2541</th>\n",
       "      <td>prom</td>\n",
       "      <td>0.281</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1987</th>\n",
       "      <td>gucci</td>\n",
       "      <td>0.275</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1783</th>\n",
       "      <td>diamond</td>\n",
       "      <td>0.247</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1906</th>\n",
       "      <td>flawless</td>\n",
       "      <td>0.237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>895</th>\n",
       "      <td>Sub_Cat_2_Athletic</td>\n",
       "      <td>0.237</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1003</th>\n",
       "      <td>Sub_Cat_2_Consoles</td>\n",
       "      <td>0.236</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2534</th>\n",
       "      <td>priority</td>\n",
       "      <td>0.234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1132</th>\n",
       "      <td>Sub_Cat_2_Laptops &amp; Netbooks</td>\n",
       "      <td>0.234</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3051</th>\n",
       "      <td>watch</td>\n",
       "      <td>0.231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1823</th>\n",
       "      <td>dust bag</td>\n",
       "      <td>0.228</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>875</th>\n",
       "      <td>Sub_Cat_1_Women's Handbags</td>\n",
       "      <td>0.214</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1020</th>\n",
       "      <td>Sub_Cat_2_Digital Cameras</td>\n",
       "      <td>0.211</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2295</th>\n",
       "      <td>mk</td>\n",
       "      <td>0.210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1695</th>\n",
       "      <td>comes</td>\n",
       "      <td>0.207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2659</th>\n",
       "      <td>sarah</td>\n",
       "      <td>0.207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3018</th>\n",
       "      <td>value</td>\n",
       "      <td>0.207</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1157</th>\n",
       "      <td>Sub_Cat_2_Messenger &amp; Crossbody</td>\n",
       "      <td>0.205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1389</th>\n",
       "      <td>30</td>\n",
       "      <td>0.201</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2586</th>\n",
       "      <td>receipt</td>\n",
       "      <td>0.194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2600</th>\n",
       "      <td>reserved</td>\n",
       "      <td>0.191</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1679</th>\n",
       "      <td>coat</td>\n",
       "      <td>0.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2516</th>\n",
       "      <td>power</td>\n",
       "      <td>0.190</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2111</th>\n",
       "      <td>kendra</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2112</th>\n",
       "      <td>kendra scott</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1556</th>\n",
       "      <td>box</td>\n",
       "      <td>0.188</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1624</th>\n",
       "      <td>carrier</td>\n",
       "      <td>0.186</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1585</th>\n",
       "      <td>bundle</td>\n",
       "      <td>0.184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2653</th>\n",
       "      <td>salt</td>\n",
       "      <td>0.183</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2347</th>\n",
       "      <td>new tag</td>\n",
       "      <td>0.182</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1538</th>\n",
       "      <td>bnwt</td>\n",
       "      <td>0.181</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3046</th>\n",
       "      <td>wallet</td>\n",
       "      <td>0.178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2676</th>\n",
       "      <td>screen</td>\n",
       "      <td>0.178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1321</th>\n",
       "      <td>Sub_Cat_2_Vacuums &amp; Floor Care</td>\n",
       "      <td>0.176</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1640</th>\n",
       "      <td>charger</td>\n",
       "      <td>0.173</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1112</th>\n",
       "      <td>Sub_Cat_2_Jackets</td>\n",
       "      <td>0.172</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             Feature_name  Ridge\n",
       "976   Sub_Cat_2_Cell Phones & Smartphones  0.535\n",
       "2228                            lululemon  0.442\n",
       "2223                              lularoe  0.393\n",
       "851                       Sub_Cat_1_Shoes  0.349\n",
       "1479                            authentic  0.312\n",
       "2541                                 prom  0.281\n",
       "1987                                gucci  0.275\n",
       "1783                              diamond  0.247\n",
       "1906                             flawless  0.237\n",
       "895                    Sub_Cat_2_Athletic  0.237\n",
       "1003                   Sub_Cat_2_Consoles  0.236\n",
       "2534                             priority  0.234\n",
       "1132         Sub_Cat_2_Laptops & Netbooks  0.234\n",
       "3051                                watch  0.231\n",
       "1823                             dust bag  0.228\n",
       "875            Sub_Cat_1_Women's Handbags  0.214\n",
       "1020            Sub_Cat_2_Digital Cameras  0.211\n",
       "2295                                   mk  0.210\n",
       "1695                                comes  0.207\n",
       "2659                                sarah  0.207\n",
       "3018                                value  0.207\n",
       "1157      Sub_Cat_2_Messenger & Crossbody  0.205\n",
       "1389                                   30  0.201\n",
       "2586                              receipt  0.194\n",
       "2600                             reserved  0.191\n",
       "1679                                 coat  0.190\n",
       "2516                                power  0.190\n",
       "2111                               kendra  0.188\n",
       "2112                         kendra scott  0.188\n",
       "1556                                  box  0.188\n",
       "1624                              carrier  0.186\n",
       "1585                               bundle  0.184\n",
       "2653                                 salt  0.183\n",
       "2347                              new tag  0.182\n",
       "1538                                 bnwt  0.181\n",
       "3046                               wallet  0.178\n",
       "2676                               screen  0.178\n",
       "1321       Sub_Cat_2_Vacuums & Floor Care  0.176\n",
       "1640                              charger  0.173\n",
       "1112                    Sub_Cat_2_Jackets  0.172"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "coeffs[~coeffs[\"Feature_name\"].str.contains(\"Brand\")].head(40)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "coeffs_head = coeffs.head(10)\n",
    "plt.figure(figsize=(16,10))\n",
    "sns.barplot(y=\"Feature_name\", x=\"Ridge\", data=coeffs_head, palette=\"viridis\")\n",
    "sns.color_palette(\"Paired\")\n",
    "plt.xlabel(\"Ridge Coefficient\", fontsize=20)\n",
    "plt.ylabel(\"Feature Name\", fontsize=20)\n",
    "plt.title(\"Feature Importances\", fontsize=30)\n",
    "plt.xticks(rotation =45)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "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>Feature_name</th>\n",
       "      <th>Ridge</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>2242</th>\n",
       "      <td>makeup</td>\n",
       "      <td>-0.151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1213</th>\n",
       "      <td>Sub_Cat_2_Pumps</td>\n",
       "      <td>-0.151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2490</th>\n",
       "      <td>plug</td>\n",
       "      <td>-0.151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1588</th>\n",
       "      <td>bundle deals</td>\n",
       "      <td>-0.151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2576</th>\n",
       "      <td>ratings</td>\n",
       "      <td>-0.152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2133</th>\n",
       "      <td>lanyard</td>\n",
       "      <td>-0.152</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2619</th>\n",
       "      <td>rings</td>\n",
       "      <td>-0.153</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2654</th>\n",
       "      <td>sample</td>\n",
       "      <td>-0.154</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1669</th>\n",
       "      <td>clear</td>\n",
       "      <td>-0.155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2369</th>\n",
       "      <td>nose</td>\n",
       "      <td>-0.156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1659</th>\n",
       "      <td>choker</td>\n",
       "      <td>-0.157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>476</th>\n",
       "      <td>Brand_Motherhood Maternity</td>\n",
       "      <td>-0.157</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1035</th>\n",
       "      <td>Sub_Cat_2_Earrings</td>\n",
       "      <td>-0.158</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1043</th>\n",
       "      <td>Sub_Cat_2_Eyes</td>\n",
       "      <td>-0.159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1945</th>\n",
       "      <td>game</td>\n",
       "      <td>-0.159</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2000</th>\n",
       "      <td>happy</td>\n",
       "      <td>-0.160</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>975</th>\n",
       "      <td>Sub_Cat_2_Cell Phone Accessories</td>\n",
       "      <td>-0.161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1257</th>\n",
       "      <td>Sub_Cat_2_Socks</td>\n",
       "      <td>-0.161</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1233</th>\n",
       "      <td>Sub_Cat_2_Screen Protectors</td>\n",
       "      <td>-0.162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2784</th>\n",
       "      <td>size xl</td>\n",
       "      <td>-0.167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1068</th>\n",
       "      <td>Sub_Cat_2_Games</td>\n",
       "      <td>-0.167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1769</th>\n",
       "      <td>decor</td>\n",
       "      <td>-0.168</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1079</th>\n",
       "      <td>Sub_Cat_2_Hair Accessories</td>\n",
       "      <td>-0.169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>955</th>\n",
       "      <td>Sub_Cat_2_Cables &amp; Adapters</td>\n",
       "      <td>-0.169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1835</th>\n",
       "      <td>elastic</td>\n",
       "      <td>-0.170</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2801</th>\n",
       "      <td>slime</td>\n",
       "      <td>-0.172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>871</th>\n",
       "      <td>Sub_Cat_1_Underwear</td>\n",
       "      <td>-0.172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1288</th>\n",
       "      <td>Sub_Cat_2_Tank, Cami</td>\n",
       "      <td>-0.172</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1260</th>\n",
       "      <td>Sub_Cat_2_Sports</td>\n",
       "      <td>-0.177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2109</th>\n",
       "      <td>kawaii</td>\n",
       "      <td>-0.178</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1267</th>\n",
       "      <td>Sub_Cat_2_Sticker</td>\n",
       "      <td>-0.180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>844</th>\n",
       "      <td>Sub_Cat_1_Paper Goods</td>\n",
       "      <td>-0.180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>516</th>\n",
       "      <td>Brand_Old Navy</td>\n",
       "      <td>-0.180</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1591</th>\n",
       "      <td>bundle save</td>\n",
       "      <td>-0.184</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>832</th>\n",
       "      <td>Sub_Cat_1_Media</td>\n",
       "      <td>-0.187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3070</th>\n",
       "      <td>wii</td>\n",
       "      <td>-0.194</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2967</th>\n",
       "      <td>tracking</td>\n",
       "      <td>-0.196</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1698</th>\n",
       "      <td>comes smoke</td>\n",
       "      <td>-0.198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2856</th>\n",
       "      <td>sticker</td>\n",
       "      <td>-0.198</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>972</th>\n",
       "      <td>Sub_Cat_2_Cases, Covers &amp; Skins</td>\n",
       "      <td>-0.206</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1198</th>\n",
       "      <td>Sub_Cat_2_Personal Care</td>\n",
       "      <td>-0.210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>793</th>\n",
       "      <td>Sub_Cat_1_Cell Phones &amp; Accessories</td>\n",
       "      <td>-0.216</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1838</th>\n",
       "      <td>envelope</td>\n",
       "      <td>-0.220</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>242</th>\n",
       "      <td>Brand_FOREVER 21</td>\n",
       "      <td>-0.240</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>shipping</td>\n",
       "      <td>-0.246</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1148</th>\n",
       "      <td>Sub_Cat_2_Makeup Brushes &amp; Tools</td>\n",
       "      <td>-0.249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1007</th>\n",
       "      <td>Sub_Cat_2_Cosmetic Bags</td>\n",
       "      <td>-0.255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>870</th>\n",
       "      <td>Sub_Cat_1_Trading Cards</td>\n",
       "      <td>-0.264</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>item_condition_id</td>\n",
       "      <td>-0.294</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2857</th>\n",
       "      <td>stickers</td>\n",
       "      <td>-0.315</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                             Feature_name  Ridge\n",
       "2242                               makeup -0.151\n",
       "1213                      Sub_Cat_2_Pumps -0.151\n",
       "2490                                 plug -0.151\n",
       "1588                         bundle deals -0.151\n",
       "2576                              ratings -0.152\n",
       "2133                              lanyard -0.152\n",
       "2619                                rings -0.153\n",
       "2654                               sample -0.154\n",
       "1669                                clear -0.155\n",
       "2369                                 nose -0.156\n",
       "1659                               choker -0.157\n",
       "476            Brand_Motherhood Maternity -0.157\n",
       "1035                   Sub_Cat_2_Earrings -0.158\n",
       "1043                       Sub_Cat_2_Eyes -0.159\n",
       "1945                                 game -0.159\n",
       "2000                                happy -0.160\n",
       "975      Sub_Cat_2_Cell Phone Accessories -0.161\n",
       "1257                      Sub_Cat_2_Socks -0.161\n",
       "1233          Sub_Cat_2_Screen Protectors -0.162\n",
       "2784                              size xl -0.167\n",
       "1068                      Sub_Cat_2_Games -0.167\n",
       "1769                                decor -0.168\n",
       "1079           Sub_Cat_2_Hair Accessories -0.169\n",
       "955           Sub_Cat_2_Cables & Adapters -0.169\n",
       "1835                              elastic -0.170\n",
       "2801                                slime -0.172\n",
       "871                   Sub_Cat_1_Underwear -0.172\n",
       "1288                 Sub_Cat_2_Tank, Cami -0.172\n",
       "1260                     Sub_Cat_2_Sports -0.177\n",
       "2109                               kawaii -0.178\n",
       "1267                    Sub_Cat_2_Sticker -0.180\n",
       "844                 Sub_Cat_1_Paper Goods -0.180\n",
       "516                        Brand_Old Navy -0.180\n",
       "1591                          bundle save -0.184\n",
       "832                       Sub_Cat_1_Media -0.187\n",
       "3070                                  wii -0.194\n",
       "2967                             tracking -0.196\n",
       "1698                          comes smoke -0.198\n",
       "2856                              sticker -0.198\n",
       "972       Sub_Cat_2_Cases, Covers & Skins -0.206\n",
       "1198              Sub_Cat_2_Personal Care -0.210\n",
       "793   Sub_Cat_1_Cell Phones & Accessories -0.216\n",
       "1838                             envelope -0.220\n",
       "242                      Brand_FOREVER 21 -0.240\n",
       "1                                shipping -0.246\n",
       "1148     Sub_Cat_2_Makeup Brushes & Tools -0.249\n",
       "1007              Sub_Cat_2_Cosmetic Bags -0.255\n",
       "870               Sub_Cat_1_Trading Cards -0.264\n",
       "0                       item_condition_id -0.294\n",
       "2857                             stickers -0.315"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "display(coeffs.tail(50))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "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>name</th>\n",
       "      <th>unprocessed_text</th>\n",
       "      <th>KNNRegressor</th>\n",
       "      <th>RandomForestRegressor</th>\n",
       "      <th>Ridge</th>\n",
       "      <th>Lasso</th>\n",
       "      <th>Linear Regression</th>\n",
       "      <th>XG Boost</th>\n",
       "      <th>Light GBM</th>\n",
       "      <th>Actual Price</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>5780</th>\n",
       "      <td>Nike lebron 13 usa olympics size 12</td>\n",
       "      <td>nike lebron 13 usa olympics size 12 here is a ...</td>\n",
       "      <td>34.56</td>\n",
       "      <td>51.54</td>\n",
       "      <td>181.45</td>\n",
       "      <td>136.90</td>\n",
       "      <td>688.90</td>\n",
       "      <td>42.56</td>\n",
       "      <td>103.16</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>337</th>\n",
       "      <td>NEW Coach Ray Outline Sig High Tops</td>\n",
       "      <td>new coach ray outline sig high tops coach ray ...</td>\n",
       "      <td>17.34</td>\n",
       "      <td>77.67</td>\n",
       "      <td>140.97</td>\n",
       "      <td>84.69</td>\n",
       "      <td>261.06</td>\n",
       "      <td>52.99</td>\n",
       "      <td>116.79</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8687</th>\n",
       "      <td>(24hrs sale) Macbook air 11.6in (2013)</td>\n",
       "      <td>(24hrs sale) macbook air 11.6in (2013) ❌❌firm ...</td>\n",
       "      <td>18.50</td>\n",
       "      <td>28.58</td>\n",
       "      <td>139.05</td>\n",
       "      <td>167.95</td>\n",
       "      <td>1458.71</td>\n",
       "      <td>96.13</td>\n",
       "      <td>76.23</td>\n",
       "      <td>550.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4961</th>\n",
       "      <td>Jordan Retro 11 Low UNC</td>\n",
       "      <td>jordan retro 11 low unc trying to sell asap. t...</td>\n",
       "      <td>50.09</td>\n",
       "      <td>102.80</td>\n",
       "      <td>124.76</td>\n",
       "      <td>89.77</td>\n",
       "      <td>227.38</td>\n",
       "      <td>94.88</td>\n",
       "      <td>114.45</td>\n",
       "      <td>109.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4656</th>\n",
       "      <td>NMD R_1 White/Black Reflective M10.5US</td>\n",
       "      <td>nmd r_1 white/black reflective m10.5us ****men...</td>\n",
       "      <td>49.12</td>\n",
       "      <td>106.51</td>\n",
       "      <td>119.00</td>\n",
       "      <td>91.78</td>\n",
       "      <td>169.39</td>\n",
       "      <td>99.01</td>\n",
       "      <td>103.67</td>\n",
       "      <td>186.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>149</th>\n",
       "      <td>LIKE NEW IPHONE 5C</td>\n",
       "      <td>like new iphone 5c just upgraded so now finall...</td>\n",
       "      <td>17.61</td>\n",
       "      <td>147.25</td>\n",
       "      <td>111.76</td>\n",
       "      <td>108.76</td>\n",
       "      <td>169.79</td>\n",
       "      <td>126.87</td>\n",
       "      <td>134.30</td>\n",
       "      <td>104.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9340</th>\n",
       "      <td>Kate Spade Sunglasses</td>\n",
       "      <td>kate spade sunglasses authentic gold and pink ...</td>\n",
       "      <td>27.90</td>\n",
       "      <td>39.68</td>\n",
       "      <td>107.63</td>\n",
       "      <td>82.77</td>\n",
       "      <td>59.26</td>\n",
       "      <td>40.14</td>\n",
       "      <td>76.24</td>\n",
       "      <td>56.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7381</th>\n",
       "      <td>Air Jordan 11 Low Concord</td>\n",
       "      <td>air jordan 11 low concord great condition 8/10...</td>\n",
       "      <td>54.42</td>\n",
       "      <td>77.21</td>\n",
       "      <td>103.57</td>\n",
       "      <td>97.16</td>\n",
       "      <td>221.61</td>\n",
       "      <td>60.64</td>\n",
       "      <td>84.71</td>\n",
       "      <td>151.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7949</th>\n",
       "      <td>Acer aspire switch New</td>\n",
       "      <td>acer aspire switch new acer aspire switch 10 s...</td>\n",
       "      <td>27.26</td>\n",
       "      <td>82.73</td>\n",
       "      <td>101.76</td>\n",
       "      <td>150.52</td>\n",
       "      <td>786.36</td>\n",
       "      <td>75.72</td>\n",
       "      <td>66.16</td>\n",
       "      <td>81.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5234</th>\n",
       "      <td>Apple Watch Series 1 38mm</td>\n",
       "      <td>apple watch series 1 38mm in excellent conditi...</td>\n",
       "      <td>26.14</td>\n",
       "      <td>128.81</td>\n",
       "      <td>101.56</td>\n",
       "      <td>161.95</td>\n",
       "      <td>115.81</td>\n",
       "      <td>167.76</td>\n",
       "      <td>181.10</td>\n",
       "      <td>246.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                            name  \\\n",
       "train_id                                           \n",
       "5780         Nike lebron 13 usa olympics size 12   \n",
       "337          NEW Coach Ray Outline Sig High Tops   \n",
       "8687      (24hrs sale) Macbook air 11.6in (2013)   \n",
       "4961                     Jordan Retro 11 Low UNC   \n",
       "4656      NMD R_1 White/Black Reflective M10.5US   \n",
       "149                           LIKE NEW IPHONE 5C   \n",
       "9340                       Kate Spade Sunglasses   \n",
       "7381                   Air Jordan 11 Low Concord   \n",
       "7949                      Acer aspire switch New   \n",
       "5234                   Apple Watch Series 1 38mm   \n",
       "\n",
       "                                           unprocessed_text  KNNRegressor  \\\n",
       "train_id                                                                    \n",
       "5780      nike lebron 13 usa olympics size 12 here is a ...         34.56   \n",
       "337       new coach ray outline sig high tops coach ray ...         17.34   \n",
       "8687      (24hrs sale) macbook air 11.6in (2013) ❌❌firm ...         18.50   \n",
       "4961      jordan retro 11 low unc trying to sell asap. t...         50.09   \n",
       "4656      nmd r_1 white/black reflective m10.5us ****men...         49.12   \n",
       "149       like new iphone 5c just upgraded so now finall...         17.61   \n",
       "9340      kate spade sunglasses authentic gold and pink ...         27.90   \n",
       "7381      air jordan 11 low concord great condition 8/10...         54.42   \n",
       "7949      acer aspire switch new acer aspire switch 10 s...         27.26   \n",
       "5234      apple watch series 1 38mm in excellent conditi...         26.14   \n",
       "\n",
       "          RandomForestRegressor   Ridge   Lasso  Linear Regression  XG Boost  \\\n",
       "train_id                                                                       \n",
       "5780                      51.54  181.45  136.90             688.90     42.56   \n",
       "337                       77.67  140.97   84.69             261.06     52.99   \n",
       "8687                      28.58  139.05  167.95            1458.71     96.13   \n",
       "4961                     102.80  124.76   89.77             227.38     94.88   \n",
       "4656                     106.51  119.00   91.78             169.39     99.01   \n",
       "149                      147.25  111.76  108.76             169.79    126.87   \n",
       "9340                      39.68  107.63   82.77              59.26     40.14   \n",
       "7381                      77.21  103.57   97.16             221.61     60.64   \n",
       "7949                      82.73  101.76  150.52             786.36     75.72   \n",
       "5234                     128.81  101.56  161.95             115.81    167.76   \n",
       "\n",
       "          Light GBM  Actual Price  \n",
       "train_id                           \n",
       "5780         103.16          56.0  \n",
       "337          116.79          56.0  \n",
       "8687          76.23         550.0  \n",
       "4961         114.45         109.0  \n",
       "4656         103.67         186.0  \n",
       "149          134.30         104.0  \n",
       "9340          76.24          56.0  \n",
       "7381          84.71         151.0  \n",
       "7949          66.16          81.0  \n",
       "5234         181.10         246.0  "
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.sort_values(by=[\"Ridge\"], ascending=False).head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "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>name</th>\n",
       "      <th>item_condition_id</th>\n",
       "      <th>brand_name</th>\n",
       "      <th>price</th>\n",
       "      <th>shipping</th>\n",
       "      <th>item_description</th>\n",
       "      <th>general_cat</th>\n",
       "      <th>subcat_1</th>\n",
       "      <th>subcat_2</th>\n",
       "      <th>unprocessed_text</th>\n",
       "      <th>processed_text</th>\n",
       "      <th>lemmatized_text</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "Empty DataFrame\n",
       "Columns: [name, item_condition_id, brand_name, price, shipping, item_description, general_cat, subcat_1, subcat_2, unprocessed_text, processed_text, lemmatized_text]\n",
       "Index: []"
      ]
     },
     "execution_count": 56,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df[df[\"name\"].str.contains(\"Rolex\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Lasso(alpha=1.0, copy_X=True, fit_intercept=True, max_iter=1000,\n",
       "   normalize=False, positive=False, precompute=False, random_state=None,\n",
       "   selection='cyclic', tol=0.0001, warm_start=False)"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr_ensemble = Lasso()\n",
    "X_ = result[[\"Ridge\", \"Lasso\",\"Linear Regression\",\"XG Boost\", \"Light GBM\"]]\n",
    "y_ = result[[\"Actual Price\"]]\n",
    "lr_ensemble.fit(X_,y_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-5.08037244])"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr_ensemble.intercept_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([-0.03744822,  0.6981428 ,  0.12617157,  0.44975318,  0.28882805])"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lr_ensemble.coef_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "result[\"ensemble\"] = lr_ensemble.predict(X_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "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>name</th>\n",
       "      <th>unprocessed_text</th>\n",
       "      <th>KNNRegressor</th>\n",
       "      <th>RandomForestRegressor</th>\n",
       "      <th>Ridge</th>\n",
       "      <th>Lasso</th>\n",
       "      <th>Linear Regression</th>\n",
       "      <th>XG Boost</th>\n",
       "      <th>Light GBM</th>\n",
       "      <th>Actual Price</th>\n",
       "      <th>ensemble</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>train_id</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6252</th>\n",
       "      <td>N64 super mario 64 (label stain)</td>\n",
       "      <td>n64 super mario 64 (label stain) great item gr...</td>\n",
       "      <td>17.72</td>\n",
       "      <td>14.87</td>\n",
       "      <td>9.92</td>\n",
       "      <td>10.68</td>\n",
       "      <td>14.23</td>\n",
       "      <td>15.98</td>\n",
       "      <td>13.40</td>\n",
       "      <td>21.0</td>\n",
       "      <td>14.857079</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4684</th>\n",
       "      <td>Jordan 9c Retro Gamma 11s</td>\n",
       "      <td>jordan 9c retro gamma 11s size 9c gamma gamma ...</td>\n",
       "      <td>19.74</td>\n",
       "      <td>24.65</td>\n",
       "      <td>24.07</td>\n",
       "      <td>21.92</td>\n",
       "      <td>35.60</td>\n",
       "      <td>17.86</td>\n",
       "      <td>21.67</td>\n",
       "      <td>29.0</td>\n",
       "      <td>28.104742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1731</th>\n",
       "      <td>5 Pair of Lashes + Lash Tweezers</td>\n",
       "      <td>5 pair of lashes + lash tweezers perfect for t...</td>\n",
       "      <td>11.24</td>\n",
       "      <td>11.39</td>\n",
       "      <td>5.30</td>\n",
       "      <td>7.59</td>\n",
       "      <td>5.94</td>\n",
       "      <td>13.08</td>\n",
       "      <td>11.67</td>\n",
       "      <td>7.0</td>\n",
       "      <td>10.022910</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4742</th>\n",
       "      <td>Bundle for LizKay</td>\n",
       "      <td>bundle for lizkay gb girls size 10. fits like ...</td>\n",
       "      <td>15.44</td>\n",
       "      <td>19.64</td>\n",
       "      <td>22.27</td>\n",
       "      <td>21.16</td>\n",
       "      <td>25.10</td>\n",
       "      <td>19.21</td>\n",
       "      <td>19.22</td>\n",
       "      <td>40.0</td>\n",
       "      <td>26.216297</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4521</th>\n",
       "      <td>New Balance Minimus size 8 - women</td>\n",
       "      <td>new balance minimus size 8 - women good used c...</td>\n",
       "      <td>24.32</td>\n",
       "      <td>26.33</td>\n",
       "      <td>34.00</td>\n",
       "      <td>31.79</td>\n",
       "      <td>23.03</td>\n",
       "      <td>27.76</td>\n",
       "      <td>24.85</td>\n",
       "      <td>16.0</td>\n",
       "      <td>38.408604</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6340</th>\n",
       "      <td>PINK Fanny pack</td>\n",
       "      <td>pink fanny pack never been used, pink fanny pa...</td>\n",
       "      <td>22.48</td>\n",
       "      <td>11.78</td>\n",
       "      <td>15.51</td>\n",
       "      <td>16.71</td>\n",
       "      <td>8.88</td>\n",
       "      <td>14.97</td>\n",
       "      <td>15.92</td>\n",
       "      <td>10.0</td>\n",
       "      <td>18.456123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>576</th>\n",
       "      <td>OFRA BEVERLY HILLS HIGHLIGHTER</td>\n",
       "      <td>ofra beverly hills highlighter swatched full size</td>\n",
       "      <td>13.84</td>\n",
       "      <td>11.53</td>\n",
       "      <td>13.74</td>\n",
       "      <td>14.04</td>\n",
       "      <td>16.09</td>\n",
       "      <td>13.75</td>\n",
       "      <td>14.05</td>\n",
       "      <td>19.0</td>\n",
       "      <td>16.479255</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5202</th>\n",
       "      <td>Victoria secret swim small</td>\n",
       "      <td>victoria secret swim small victoria secret swi...</td>\n",
       "      <td>15.83</td>\n",
       "      <td>15.15</td>\n",
       "      <td>16.76</td>\n",
       "      <td>16.84</td>\n",
       "      <td>21.17</td>\n",
       "      <td>17.03</td>\n",
       "      <td>15.21</td>\n",
       "      <td>10.0</td>\n",
       "      <td>20.772143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6363</th>\n",
       "      <td>MAD MATTE w/liner</td>\n",
       "      <td>mad matte w/liner liquid matte sold as a set o...</td>\n",
       "      <td>15.97</td>\n",
       "      <td>13.90</td>\n",
       "      <td>10.24</td>\n",
       "      <td>12.45</td>\n",
       "      <td>6.51</td>\n",
       "      <td>15.06</td>\n",
       "      <td>13.62</td>\n",
       "      <td>6.0</td>\n",
       "      <td>14.756533</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>439</th>\n",
       "      <td>Jaclyn Hill Palette</td>\n",
       "      <td>jaclyn hill palette brand new! authentic!! abs...</td>\n",
       "      <td>24.94</td>\n",
       "      <td>32.15</td>\n",
       "      <td>32.95</td>\n",
       "      <td>32.13</td>\n",
       "      <td>36.63</td>\n",
       "      <td>26.62</td>\n",
       "      <td>36.47</td>\n",
       "      <td>69.0</td>\n",
       "      <td>43.244690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2750</th>\n",
       "      <td>Baby girl Car seat canopy</td>\n",
       "      <td>baby girl car seat canopy in euc. inside is so...</td>\n",
       "      <td>18.50</td>\n",
       "      <td>17.51</td>\n",
       "      <td>15.08</td>\n",
       "      <td>12.97</td>\n",
       "      <td>15.97</td>\n",
       "      <td>17.73</td>\n",
       "      <td>14.43</td>\n",
       "      <td>13.0</td>\n",
       "      <td>17.566693</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7487</th>\n",
       "      <td>WOMEN'S CHAMPION PERFORMANCE L/S TEE</td>\n",
       "      <td>women's champion performance l/s tee this perf...</td>\n",
       "      <td>14.41</td>\n",
       "      <td>10.04</td>\n",
       "      <td>9.31</td>\n",
       "      <td>10.02</td>\n",
       "      <td>6.29</td>\n",
       "      <td>13.94</td>\n",
       "      <td>12.04</td>\n",
       "      <td>13.0</td>\n",
       "      <td>12.107044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5272</th>\n",
       "      <td>FREE SHIP VS PINK Boyshort</td>\n",
       "      <td>free ship vs pink boyshort victoria's secret p...</td>\n",
       "      <td>14.23</td>\n",
       "      <td>18.52</td>\n",
       "      <td>15.81</td>\n",
       "      <td>15.39</td>\n",
       "      <td>15.70</td>\n",
       "      <td>15.59</td>\n",
       "      <td>17.71</td>\n",
       "      <td>10.0</td>\n",
       "      <td>19.179679</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5653</th>\n",
       "      <td>HOLD Bundle ABERCROMBIE Jeans 14 Stella</td>\n",
       "      <td>hold bundle abercrombie jeans 14 stella bundle...</td>\n",
       "      <td>13.88</td>\n",
       "      <td>16.30</td>\n",
       "      <td>19.46</td>\n",
       "      <td>17.08</td>\n",
       "      <td>27.00</td>\n",
       "      <td>13.35</td>\n",
       "      <td>16.65</td>\n",
       "      <td>25.0</td>\n",
       "      <td>20.334989</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3999</th>\n",
       "      <td>Orlando Magic Shaquille O'Neal Jersey</td>\n",
       "      <td>orlando magic shaquille o'neal jersey nwt nike...</td>\n",
       "      <td>20.25</td>\n",
       "      <td>29.32</td>\n",
       "      <td>22.17</td>\n",
       "      <td>20.04</td>\n",
       "      <td>33.67</td>\n",
       "      <td>13.71</td>\n",
       "      <td>18.31</td>\n",
       "      <td>40.0</td>\n",
       "      <td>23.782937</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6033</th>\n",
       "      <td>TWO Juicy Couture bracelets</td>\n",
       "      <td>two juicy couture bracelets two beautiful juic...</td>\n",
       "      <td>14.55</td>\n",
       "      <td>27.64</td>\n",
       "      <td>28.76</td>\n",
       "      <td>30.73</td>\n",
       "      <td>21.14</td>\n",
       "      <td>26.44</td>\n",
       "      <td>32.08</td>\n",
       "      <td>22.0</td>\n",
       "      <td>39.120890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>582</th>\n",
       "      <td>iPhone 6/6s plus cases</td>\n",
       "      <td>iphone 6/6s plus cases 4 iphone 6/6s plus cases.</td>\n",
       "      <td>8.47</td>\n",
       "      <td>11.53</td>\n",
       "      <td>12.81</td>\n",
       "      <td>12.84</td>\n",
       "      <td>12.91</td>\n",
       "      <td>11.77</td>\n",
       "      <td>11.55</td>\n",
       "      <td>20.0</td>\n",
       "      <td>13.662503</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9930</th>\n",
       "      <td>MÉDIUM BLACK WAIST TRAINER CORSET</td>\n",
       "      <td>médium black waist trainer corset top quality!...</td>\n",
       "      <td>27.14</td>\n",
       "      <td>14.32</td>\n",
       "      <td>13.95</td>\n",
       "      <td>11.23</td>\n",
       "      <td>15.27</td>\n",
       "      <td>14.05</td>\n",
       "      <td>13.32</td>\n",
       "      <td>11.0</td>\n",
       "      <td>14.330230</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7051</th>\n",
       "      <td>Polo Boots</td>\n",
       "      <td>polo boots</td>\n",
       "      <td>22.12</td>\n",
       "      <td>15.43</td>\n",
       "      <td>20.39</td>\n",
       "      <td>19.38</td>\n",
       "      <td>17.35</td>\n",
       "      <td>17.97</td>\n",
       "      <td>16.97</td>\n",
       "      <td>56.0</td>\n",
       "      <td>22.858619</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8158</th>\n",
       "      <td>5 pairs baby boy shoes lot</td>\n",
       "      <td>5 pairs baby boy shoes lot 5 pairs of baby boy...</td>\n",
       "      <td>11.90</td>\n",
       "      <td>12.52</td>\n",
       "      <td>16.22</td>\n",
       "      <td>15.51</td>\n",
       "      <td>14.17</td>\n",
       "      <td>13.35</td>\n",
       "      <td>14.78</td>\n",
       "      <td>10.0</td>\n",
       "      <td>17.201347</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9896</th>\n",
       "      <td>Men's New Balance 574 Green/White Sz 9M</td>\n",
       "      <td>men's new balance 574 green/white sz 9m mens n...</td>\n",
       "      <td>34.46</td>\n",
       "      <td>22.41</td>\n",
       "      <td>30.78</td>\n",
       "      <td>33.37</td>\n",
       "      <td>20.06</td>\n",
       "      <td>31.69</td>\n",
       "      <td>29.00</td>\n",
       "      <td>26.0</td>\n",
       "      <td>42.223690</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2249</th>\n",
       "      <td>Cute authentic Hawaiian backpack</td>\n",
       "      <td>cute authentic hawaiian backpack made in hawai...</td>\n",
       "      <td>13.99</td>\n",
       "      <td>22.90</td>\n",
       "      <td>25.84</td>\n",
       "      <td>24.02</td>\n",
       "      <td>24.67</td>\n",
       "      <td>25.75</td>\n",
       "      <td>23.62</td>\n",
       "      <td>13.0</td>\n",
       "      <td>32.237271</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4640</th>\n",
       "      <td>SALE Kendall &amp; Kylie Denim Biker Jacket</td>\n",
       "      <td>sale kendall &amp; kylie denim biker jacket limite...</td>\n",
       "      <td>21.74</td>\n",
       "      <td>25.68</td>\n",
       "      <td>19.05</td>\n",
       "      <td>21.58</td>\n",
       "      <td>12.02</td>\n",
       "      <td>21.74</td>\n",
       "      <td>22.28</td>\n",
       "      <td>16.0</td>\n",
       "      <td>27.001466</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9485</th>\n",
       "      <td>Creative co-op creamer</td>\n",
       "      <td>creative co-op creamer compliments your rae dunn</td>\n",
       "      <td>18.67</td>\n",
       "      <td>17.69</td>\n",
       "      <td>20.03</td>\n",
       "      <td>19.54</td>\n",
       "      <td>16.41</td>\n",
       "      <td>18.86</td>\n",
       "      <td>20.33</td>\n",
       "      <td>12.0</td>\n",
       "      <td>24.235945</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4947</th>\n",
       "      <td>S/M Lularoe Toystory</td>\n",
       "      <td>s/m lularoe toystory brand new never worn</td>\n",
       "      <td>19.69</td>\n",
       "      <td>36.89</td>\n",
       "      <td>32.01</td>\n",
       "      <td>34.11</td>\n",
       "      <td>33.82</td>\n",
       "      <td>27.72</td>\n",
       "      <td>32.90</td>\n",
       "      <td>36.0</td>\n",
       "      <td>43.771284</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9920</th>\n",
       "      <td>3T Boy Pants Bundle</td>\n",
       "      <td>3t boy pants bundle 7 piece~excellent conditio...</td>\n",
       "      <td>17.59</td>\n",
       "      <td>15.08</td>\n",
       "      <td>20.24</td>\n",
       "      <td>18.69</td>\n",
       "      <td>14.97</td>\n",
       "      <td>18.20</td>\n",
       "      <td>16.01</td>\n",
       "      <td>21.0</td>\n",
       "      <td>21.908398</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1963</th>\n",
       "      <td>Becca Backlight Priming Filter 1oz</td>\n",
       "      <td>becca backlight priming filter 1oz authentic 8...</td>\n",
       "      <td>14.69</td>\n",
       "      <td>18.31</td>\n",
       "      <td>25.97</td>\n",
       "      <td>25.97</td>\n",
       "      <td>22.33</td>\n",
       "      <td>19.94</td>\n",
       "      <td>23.39</td>\n",
       "      <td>20.0</td>\n",
       "      <td>30.619043</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8243</th>\n",
       "      <td>It highlight brush</td>\n",
       "      <td>it highlight brush used once.</td>\n",
       "      <td>11.66</td>\n",
       "      <td>14.18</td>\n",
       "      <td>11.24</td>\n",
       "      <td>10.80</td>\n",
       "      <td>5.51</td>\n",
       "      <td>15.89</td>\n",
       "      <td>11.53</td>\n",
       "      <td>11.0</td>\n",
       "      <td>13.210623</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6590</th>\n",
       "      <td>USA Navy Blue Men's Jersey</td>\n",
       "      <td>usa navy blue men's jersey new jersey size med...</td>\n",
       "      <td>21.19</td>\n",
       "      <td>26.01</td>\n",
       "      <td>20.40</td>\n",
       "      <td>17.70</td>\n",
       "      <td>26.22</td>\n",
       "      <td>14.05</td>\n",
       "      <td>16.35</td>\n",
       "      <td>10.0</td>\n",
       "      <td>20.862401</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8847</th>\n",
       "      <td>Nike headphones. ( black)</td>\n",
       "      <td>nike headphones. ( black) brand new, great bas...</td>\n",
       "      <td>21.37</td>\n",
       "      <td>12.97</td>\n",
       "      <td>16.16</td>\n",
       "      <td>16.37</td>\n",
       "      <td>10.80</td>\n",
       "      <td>14.63</td>\n",
       "      <td>16.93</td>\n",
       "      <td>6.0</td>\n",
       "      <td>18.575463</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>321</th>\n",
       "      <td>Vera Bradley eyeglass case</td>\n",
       "      <td>vera bradley eyeglass case vera bradley eyegla...</td>\n",
       "      <td>16.72</td>\n",
       "      <td>15.27</td>\n",
       "      <td>15.27</td>\n",
       "      <td>17.38</td>\n",
       "      <td>14.79</td>\n",
       "      <td>18.88</td>\n",
       "      <td>15.19</td>\n",
       "      <td>10.0</td>\n",
       "      <td>21.226231</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2678</th>\n",
       "      <td>2Cargo short brand new style 2017</td>\n",
       "      <td>2cargo short brand new style 2017 cargo shorts...</td>\n",
       "      <td>15.41</td>\n",
       "      <td>17.61</td>\n",
       "      <td>36.54</td>\n",
       "      <td>23.24</td>\n",
       "      <td>27.62</td>\n",
       "      <td>20.41</td>\n",
       "      <td>23.34</td>\n",
       "      <td>56.0</td>\n",
       "      <td>29.181676</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4625</th>\n",
       "      <td>Rae Dunn Good Morning Bowl</td>\n",
       "      <td>rae dunn good morning bowl brand new. authenti...</td>\n",
       "      <td>19.54</td>\n",
       "      <td>36.90</td>\n",
       "      <td>47.32</td>\n",
       "      <td>39.85</td>\n",
       "      <td>39.16</td>\n",
       "      <td>33.22</td>\n",
       "      <td>42.66</td>\n",
       "      <td>18.0</td>\n",
       "      <td>53.171652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4949</th>\n",
       "      <td>Tank Top Lot</td>\n",
       "      <td>tank top lot 10 tanks- vs, nike, gap, old navy...</td>\n",
       "      <td>17.01</td>\n",
       "      <td>15.14</td>\n",
       "      <td>18.86</td>\n",
       "      <td>17.92</td>\n",
       "      <td>20.87</td>\n",
       "      <td>15.71</td>\n",
       "      <td>14.80</td>\n",
       "      <td>16.0</td>\n",
       "      <td>20.697551</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8328</th>\n",
       "      <td>Soma bra</td>\n",
       "      <td>soma bra black no show back soma bra size 34 d...</td>\n",
       "      <td>12.33</td>\n",
       "      <td>14.02</td>\n",
       "      <td>14.94</td>\n",
       "      <td>14.69</td>\n",
       "      <td>11.38</td>\n",
       "      <td>18.58</td>\n",
       "      <td>12.56</td>\n",
       "      <td>10.0</td>\n",
       "      <td>18.035796</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3337</th>\n",
       "      <td>ESTEE LAUDER KNOWING EAU DE PARFUM - 1OZ</td>\n",
       "      <td>estee lauder knowing eau de parfum - 1oz new e...</td>\n",
       "      <td>18.30</td>\n",
       "      <td>10.66</td>\n",
       "      <td>26.62</td>\n",
       "      <td>23.44</td>\n",
       "      <td>17.22</td>\n",
       "      <td>14.32</td>\n",
       "      <td>20.89</td>\n",
       "      <td>20.0</td>\n",
       "      <td>24.933981</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5589</th>\n",
       "      <td>Vans Off The Wall Like New (Steal)</td>\n",
       "      <td>vans off the wall like new (steal) worn twice ...</td>\n",
       "      <td>48.02</td>\n",
       "      <td>37.55</td>\n",
       "      <td>42.52</td>\n",
       "      <td>37.30</td>\n",
       "      <td>48.23</td>\n",
       "      <td>30.13</td>\n",
       "      <td>43.46</td>\n",
       "      <td>23.0</td>\n",
       "      <td>51.556841</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>251</th>\n",
       "      <td>ABH DipBrow Pomade</td>\n",
       "      <td>abh dipbrow pomade shade: dark brown has been ...</td>\n",
       "      <td>14.03</td>\n",
       "      <td>13.84</td>\n",
       "      <td>7.33</td>\n",
       "      <td>9.63</td>\n",
       "      <td>2.09</td>\n",
       "      <td>16.74</td>\n",
       "      <td>12.19</td>\n",
       "      <td>9.0</td>\n",
       "      <td>12.681628</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3973</th>\n",
       "      <td>Real Jade Bangle Bracelet 58 -62 MM</td>\n",
       "      <td>real jade bangle bracelet 58 -62 mm</td>\n",
       "      <td>15.98</td>\n",
       "      <td>11.74</td>\n",
       "      <td>13.31</td>\n",
       "      <td>14.04</td>\n",
       "      <td>10.71</td>\n",
       "      <td>14.07</td>\n",
       "      <td>11.74</td>\n",
       "      <td>30.0</td>\n",
       "      <td>15.293283</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6630</th>\n",
       "      <td>American Eagle Pants</td>\n",
       "      <td>american eagle pants skinny american eagle jea...</td>\n",
       "      <td>15.90</td>\n",
       "      <td>18.18</td>\n",
       "      <td>14.37</td>\n",
       "      <td>17.03</td>\n",
       "      <td>14.31</td>\n",
       "      <td>18.33</td>\n",
       "      <td>16.80</td>\n",
       "      <td>13.0</td>\n",
       "      <td>21.172671</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5547</th>\n",
       "      <td>Ralph Lauren brown leather boots</td>\n",
       "      <td>ralph lauren brown leather boots broken in wit...</td>\n",
       "      <td>26.36</td>\n",
       "      <td>20.89</td>\n",
       "      <td>25.48</td>\n",
       "      <td>25.40</td>\n",
       "      <td>28.45</td>\n",
       "      <td>22.68</td>\n",
       "      <td>27.11</td>\n",
       "      <td>28.0</td>\n",
       "      <td>33.318386</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>Black Capri Leggings w/ Flowers One size</td>\n",
       "      <td>black capri leggings w/ flowers one size one s...</td>\n",
       "      <td>29.38</td>\n",
       "      <td>11.37</td>\n",
       "      <td>11.81</td>\n",
       "      <td>11.83</td>\n",
       "      <td>12.15</td>\n",
       "      <td>14.38</td>\n",
       "      <td>11.16</td>\n",
       "      <td>13.0</td>\n",
       "      <td>13.960150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8362</th>\n",
       "      <td>Victoria's Secret PINK White Sox Tank</td>\n",
       "      <td>victoria's secret pink white sox tank this has...</td>\n",
       "      <td>20.26</td>\n",
       "      <td>19.30</td>\n",
       "      <td>13.13</td>\n",
       "      <td>11.52</td>\n",
       "      <td>15.69</td>\n",
       "      <td>10.68</td>\n",
       "      <td>12.79</td>\n",
       "      <td>12.0</td>\n",
       "      <td>12.947644</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1513</th>\n",
       "      <td>Papaya Distressed Light Demin Shorts S</td>\n",
       "      <td>papaya distressed light demin shorts s lowest....</td>\n",
       "      <td>16.87</td>\n",
       "      <td>21.20</td>\n",
       "      <td>12.06</td>\n",
       "      <td>11.52</td>\n",
       "      <td>18.91</td>\n",
       "      <td>18.67</td>\n",
       "      <td>15.05</td>\n",
       "      <td>9.0</td>\n",
       "      <td>17.640265</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9317</th>\n",
       "      <td>NEW Puma Soccer Ball Size 5 OfficialSize</td>\n",
       "      <td>new puma soccer ball size 5 officialsize puma ...</td>\n",
       "      <td>16.89</td>\n",
       "      <td>19.73</td>\n",
       "      <td>29.22</td>\n",
       "      <td>25.92</td>\n",
       "      <td>43.02</td>\n",
       "      <td>23.81</td>\n",
       "      <td>28.23</td>\n",
       "      <td>21.0</td>\n",
       "      <td>36.211392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>Victoria secret 34 c corest top</td>\n",
       "      <td>victoria secret 34 c corest top victoria secre...</td>\n",
       "      <td>17.36</td>\n",
       "      <td>14.60</td>\n",
       "      <td>12.49</td>\n",
       "      <td>11.35</td>\n",
       "      <td>10.08</td>\n",
       "      <td>16.48</td>\n",
       "      <td>11.65</td>\n",
       "      <td>10.0</td>\n",
       "      <td>14.424409</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4819</th>\n",
       "      <td>JujuBe Royal Envy Quick</td>\n",
       "      <td>jujube royal envy quick used good condition</td>\n",
       "      <td>12.48</td>\n",
       "      <td>14.43</td>\n",
       "      <td>11.81</td>\n",
       "      <td>11.16</td>\n",
       "      <td>16.34</td>\n",
       "      <td>13.35</td>\n",
       "      <td>11.63</td>\n",
       "      <td>10.0</td>\n",
       "      <td>13.693556</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3465</th>\n",
       "      <td>Old navy cardigan 5T</td>\n",
       "      <td>old navy cardigan 5t white cardigan old navy s...</td>\n",
       "      <td>17.14</td>\n",
       "      <td>12.17</td>\n",
       "      <td>8.38</td>\n",
       "      <td>10.03</td>\n",
       "      <td>7.13</td>\n",
       "      <td>17.73</td>\n",
       "      <td>11.27</td>\n",
       "      <td>8.0</td>\n",
       "      <td>13.737003</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1760</th>\n",
       "      <td>GAP 4T Smocked Sundress</td>\n",
       "      <td>gap 4t smocked sundress beautiful colors! worn...</td>\n",
       "      <td>13.51</td>\n",
       "      <td>11.68</td>\n",
       "      <td>13.74</td>\n",
       "      <td>13.57</td>\n",
       "      <td>15.37</td>\n",
       "      <td>13.87</td>\n",
       "      <td>12.58</td>\n",
       "      <td>11.0</td>\n",
       "      <td>15.689677</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2304</th>\n",
       "      <td>Crop top</td>\n",
       "      <td>crop top knit cropped cami</td>\n",
       "      <td>10.00</td>\n",
       "      <td>10.57</td>\n",
       "      <td>8.82</td>\n",
       "      <td>9.55</td>\n",
       "      <td>6.77</td>\n",
       "      <td>13.38</td>\n",
       "      <td>9.76</td>\n",
       "      <td>10.0</td>\n",
       "      <td>10.947439</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                              name  \\\n",
       "train_id                                             \n",
       "6252              N64 super mario 64 (label stain)   \n",
       "4684                     Jordan 9c Retro Gamma 11s   \n",
       "1731              5 Pair of Lashes + Lash Tweezers   \n",
       "4742                             Bundle for LizKay   \n",
       "4521            New Balance Minimus size 8 - women   \n",
       "6340                               PINK Fanny pack   \n",
       "576                 OFRA BEVERLY HILLS HIGHLIGHTER   \n",
       "5202                    Victoria secret swim small   \n",
       "6363                             MAD MATTE w/liner   \n",
       "439                            Jaclyn Hill Palette   \n",
       "2750                     Baby girl Car seat canopy   \n",
       "7487          WOMEN'S CHAMPION PERFORMANCE L/S TEE   \n",
       "5272                    FREE SHIP VS PINK Boyshort   \n",
       "5653       HOLD Bundle ABERCROMBIE Jeans 14 Stella   \n",
       "3999         Orlando Magic Shaquille O'Neal Jersey   \n",
       "6033                   TWO Juicy Couture bracelets   \n",
       "582                         iPhone 6/6s plus cases   \n",
       "9930             MÉDIUM BLACK WAIST TRAINER CORSET   \n",
       "7051                                    Polo Boots   \n",
       "8158                    5 pairs baby boy shoes lot   \n",
       "9896       Men's New Balance 574 Green/White Sz 9M   \n",
       "2249              Cute authentic Hawaiian backpack   \n",
       "4640       SALE Kendall & Kylie Denim Biker Jacket   \n",
       "9485                        Creative co-op creamer   \n",
       "4947                          S/M Lularoe Toystory   \n",
       "9920                           3T Boy Pants Bundle   \n",
       "1963            Becca Backlight Priming Filter 1oz   \n",
       "8243                            It highlight brush   \n",
       "6590                    USA Navy Blue Men's Jersey   \n",
       "8847                     Nike headphones. ( black)   \n",
       "321                     Vera Bradley eyeglass case   \n",
       "2678             2Cargo short brand new style 2017   \n",
       "4625                    Rae Dunn Good Morning Bowl   \n",
       "4949                                  Tank Top Lot   \n",
       "8328                                      Soma bra   \n",
       "3337      ESTEE LAUDER KNOWING EAU DE PARFUM - 1OZ   \n",
       "5589            Vans Off The Wall Like New (Steal)   \n",
       "251                             ABH DipBrow Pomade   \n",
       "3973           Real Jade Bangle Bracelet 58 -62 MM   \n",
       "6630                          American Eagle Pants   \n",
       "5547              Ralph Lauren brown leather boots   \n",
       "35        Black Capri Leggings w/ Flowers One size   \n",
       "8362         Victoria's Secret PINK White Sox Tank   \n",
       "1513        Papaya Distressed Light Demin Shorts S   \n",
       "9317      NEW Puma Soccer Ball Size 5 OfficialSize   \n",
       "39                 Victoria secret 34 c corest top   \n",
       "4819                       JujuBe Royal Envy Quick   \n",
       "3465                          Old navy cardigan 5T   \n",
       "1760                       GAP 4T Smocked Sundress   \n",
       "2304                                      Crop top   \n",
       "\n",
       "                                           unprocessed_text  KNNRegressor  \\\n",
       "train_id                                                                    \n",
       "6252      n64 super mario 64 (label stain) great item gr...         17.72   \n",
       "4684      jordan 9c retro gamma 11s size 9c gamma gamma ...         19.74   \n",
       "1731      5 pair of lashes + lash tweezers perfect for t...         11.24   \n",
       "4742      bundle for lizkay gb girls size 10. fits like ...         15.44   \n",
       "4521      new balance minimus size 8 - women good used c...         24.32   \n",
       "6340      pink fanny pack never been used, pink fanny pa...         22.48   \n",
       "576       ofra beverly hills highlighter swatched full size         13.84   \n",
       "5202      victoria secret swim small victoria secret swi...         15.83   \n",
       "6363      mad matte w/liner liquid matte sold as a set o...         15.97   \n",
       "439       jaclyn hill palette brand new! authentic!! abs...         24.94   \n",
       "2750      baby girl car seat canopy in euc. inside is so...         18.50   \n",
       "7487      women's champion performance l/s tee this perf...         14.41   \n",
       "5272      free ship vs pink boyshort victoria's secret p...         14.23   \n",
       "5653      hold bundle abercrombie jeans 14 stella bundle...         13.88   \n",
       "3999      orlando magic shaquille o'neal jersey nwt nike...         20.25   \n",
       "6033      two juicy couture bracelets two beautiful juic...         14.55   \n",
       "582        iphone 6/6s plus cases 4 iphone 6/6s plus cases.          8.47   \n",
       "9930      médium black waist trainer corset top quality!...         27.14   \n",
       "7051                                            polo boots          22.12   \n",
       "8158      5 pairs baby boy shoes lot 5 pairs of baby boy...         11.90   \n",
       "9896      men's new balance 574 green/white sz 9m mens n...         34.46   \n",
       "2249      cute authentic hawaiian backpack made in hawai...         13.99   \n",
       "4640      sale kendall & kylie denim biker jacket limite...         21.74   \n",
       "9485       creative co-op creamer compliments your rae dunn         18.67   \n",
       "4947              s/m lularoe toystory brand new never worn         19.69   \n",
       "9920      3t boy pants bundle 7 piece~excellent conditio...         17.59   \n",
       "1963      becca backlight priming filter 1oz authentic 8...         14.69   \n",
       "8243                          it highlight brush used once.         11.66   \n",
       "6590      usa navy blue men's jersey new jersey size med...         21.19   \n",
       "8847      nike headphones. ( black) brand new, great bas...         21.37   \n",
       "321       vera bradley eyeglass case vera bradley eyegla...         16.72   \n",
       "2678      2cargo short brand new style 2017 cargo shorts...         15.41   \n",
       "4625      rae dunn good morning bowl brand new. authenti...         19.54   \n",
       "4949      tank top lot 10 tanks- vs, nike, gap, old navy...         17.01   \n",
       "8328      soma bra black no show back soma bra size 34 d...         12.33   \n",
       "3337      estee lauder knowing eau de parfum - 1oz new e...         18.30   \n",
       "5589      vans off the wall like new (steal) worn twice ...         48.02   \n",
       "251       abh dipbrow pomade shade: dark brown has been ...         14.03   \n",
       "3973                   real jade bangle bracelet 58 -62 mm          15.98   \n",
       "6630      american eagle pants skinny american eagle jea...         15.90   \n",
       "5547      ralph lauren brown leather boots broken in wit...         26.36   \n",
       "35        black capri leggings w/ flowers one size one s...         29.38   \n",
       "8362      victoria's secret pink white sox tank this has...         20.26   \n",
       "1513      papaya distressed light demin shorts s lowest....         16.87   \n",
       "9317      new puma soccer ball size 5 officialsize puma ...         16.89   \n",
       "39        victoria secret 34 c corest top victoria secre...         17.36   \n",
       "4819            jujube royal envy quick used good condition         12.48   \n",
       "3465      old navy cardigan 5t white cardigan old navy s...         17.14   \n",
       "1760      gap 4t smocked sundress beautiful colors! worn...         13.51   \n",
       "2304                             crop top knit cropped cami         10.00   \n",
       "\n",
       "          RandomForestRegressor  Ridge  Lasso  Linear Regression  XG Boost  \\\n",
       "train_id                                                                     \n",
       "6252                      14.87   9.92  10.68              14.23     15.98   \n",
       "4684                      24.65  24.07  21.92              35.60     17.86   \n",
       "1731                      11.39   5.30   7.59               5.94     13.08   \n",
       "4742                      19.64  22.27  21.16              25.10     19.21   \n",
       "4521                      26.33  34.00  31.79              23.03     27.76   \n",
       "6340                      11.78  15.51  16.71               8.88     14.97   \n",
       "576                       11.53  13.74  14.04              16.09     13.75   \n",
       "5202                      15.15  16.76  16.84              21.17     17.03   \n",
       "6363                      13.90  10.24  12.45               6.51     15.06   \n",
       "439                       32.15  32.95  32.13              36.63     26.62   \n",
       "2750                      17.51  15.08  12.97              15.97     17.73   \n",
       "7487                      10.04   9.31  10.02               6.29     13.94   \n",
       "5272                      18.52  15.81  15.39              15.70     15.59   \n",
       "5653                      16.30  19.46  17.08              27.00     13.35   \n",
       "3999                      29.32  22.17  20.04              33.67     13.71   \n",
       "6033                      27.64  28.76  30.73              21.14     26.44   \n",
       "582                       11.53  12.81  12.84              12.91     11.77   \n",
       "9930                      14.32  13.95  11.23              15.27     14.05   \n",
       "7051                      15.43  20.39  19.38              17.35     17.97   \n",
       "8158                      12.52  16.22  15.51              14.17     13.35   \n",
       "9896                      22.41  30.78  33.37              20.06     31.69   \n",
       "2249                      22.90  25.84  24.02              24.67     25.75   \n",
       "4640                      25.68  19.05  21.58              12.02     21.74   \n",
       "9485                      17.69  20.03  19.54              16.41     18.86   \n",
       "4947                      36.89  32.01  34.11              33.82     27.72   \n",
       "9920                      15.08  20.24  18.69              14.97     18.20   \n",
       "1963                      18.31  25.97  25.97              22.33     19.94   \n",
       "8243                      14.18  11.24  10.80               5.51     15.89   \n",
       "6590                      26.01  20.40  17.70              26.22     14.05   \n",
       "8847                      12.97  16.16  16.37              10.80     14.63   \n",
       "321                       15.27  15.27  17.38              14.79     18.88   \n",
       "2678                      17.61  36.54  23.24              27.62     20.41   \n",
       "4625                      36.90  47.32  39.85              39.16     33.22   \n",
       "4949                      15.14  18.86  17.92              20.87     15.71   \n",
       "8328                      14.02  14.94  14.69              11.38     18.58   \n",
       "3337                      10.66  26.62  23.44              17.22     14.32   \n",
       "5589                      37.55  42.52  37.30              48.23     30.13   \n",
       "251                       13.84   7.33   9.63               2.09     16.74   \n",
       "3973                      11.74  13.31  14.04              10.71     14.07   \n",
       "6630                      18.18  14.37  17.03              14.31     18.33   \n",
       "5547                      20.89  25.48  25.40              28.45     22.68   \n",
       "35                        11.37  11.81  11.83              12.15     14.38   \n",
       "8362                      19.30  13.13  11.52              15.69     10.68   \n",
       "1513                      21.20  12.06  11.52              18.91     18.67   \n",
       "9317                      19.73  29.22  25.92              43.02     23.81   \n",
       "39                        14.60  12.49  11.35              10.08     16.48   \n",
       "4819                      14.43  11.81  11.16              16.34     13.35   \n",
       "3465                      12.17   8.38  10.03               7.13     17.73   \n",
       "1760                      11.68  13.74  13.57              15.37     13.87   \n",
       "2304                      10.57   8.82   9.55               6.77     13.38   \n",
       "\n",
       "          Light GBM  Actual Price   ensemble  \n",
       "train_id                                      \n",
       "6252          13.40          21.0  14.857079  \n",
       "4684          21.67          29.0  28.104742  \n",
       "1731          11.67           7.0  10.022910  \n",
       "4742          19.22          40.0  26.216297  \n",
       "4521          24.85          16.0  38.408604  \n",
       "6340          15.92          10.0  18.456123  \n",
       "576           14.05          19.0  16.479255  \n",
       "5202          15.21          10.0  20.772143  \n",
       "6363          13.62           6.0  14.756533  \n",
       "439           36.47          69.0  43.244690  \n",
       "2750          14.43          13.0  17.566693  \n",
       "7487          12.04          13.0  12.107044  \n",
       "5272          17.71          10.0  19.179679  \n",
       "5653          16.65          25.0  20.334989  \n",
       "3999          18.31          40.0  23.782937  \n",
       "6033          32.08          22.0  39.120890  \n",
       "582           11.55          20.0  13.662503  \n",
       "9930          13.32          11.0  14.330230  \n",
       "7051          16.97          56.0  22.858619  \n",
       "8158          14.78          10.0  17.201347  \n",
       "9896          29.00          26.0  42.223690  \n",
       "2249          23.62          13.0  32.237271  \n",
       "4640          22.28          16.0  27.001466  \n",
       "9485          20.33          12.0  24.235945  \n",
       "4947          32.90          36.0  43.771284  \n",
       "9920          16.01          21.0  21.908398  \n",
       "1963          23.39          20.0  30.619043  \n",
       "8243          11.53          11.0  13.210623  \n",
       "6590          16.35          10.0  20.862401  \n",
       "8847          16.93           6.0  18.575463  \n",
       "321           15.19          10.0  21.226231  \n",
       "2678          23.34          56.0  29.181676  \n",
       "4625          42.66          18.0  53.171652  \n",
       "4949          14.80          16.0  20.697551  \n",
       "8328          12.56          10.0  18.035796  \n",
       "3337          20.89          20.0  24.933981  \n",
       "5589          43.46          23.0  51.556841  \n",
       "251           12.19           9.0  12.681628  \n",
       "3973          11.74          30.0  15.293283  \n",
       "6630          16.80          13.0  21.172671  \n",
       "5547          27.11          28.0  33.318386  \n",
       "35            11.16          13.0  13.960150  \n",
       "8362          12.79          12.0  12.947644  \n",
       "1513          15.05           9.0  17.640265  \n",
       "9317          28.23          21.0  36.211392  \n",
       "39            11.65          10.0  14.424409  \n",
       "4819          11.63          10.0  13.693556  \n",
       "3465          11.27           8.0  13.737003  \n",
       "1760          12.58          11.0  15.689677  \n",
       "2304           9.76          10.0  10.947439  "
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.head(50)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "# display_model_predictions(rf, X_train, y_train, X_test, y_test, idx_train, idx_test, \"Random Forest\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "# df[np.logical_and(df[\"name\"].str.contains(\"Louis\"),  df[\"price\"]>1 )][\"price\"].mean()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "# result_copy = result[result.columns]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "# result_copy[\"Ridge\"] = round(((result[\"Ridge\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"Lasso\"] = round(((result[\"Lasso\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"Linear Regression\"] = round(((result[\"Linear Regression\"] - result[\"Actual Price\"])/result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"XG Boost\"] = round(((result[\"XG Boost\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"Light GBM\"] = round(((result[\"Light GBM\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"Default Model\"] = round(((result[\"Default Model\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy[\"Ensemble\"] = round(((result[\"ensemble\"]-result[\"Actual Price\"]) / result[\"Actual Price\"])*100,3)\n",
    "# result_copy.head(1000)\n",
    "\n",
    "# columns=[\"Ridge\",\"Lasso\",\"Linear Regression\",\"XG Boost\", \"Light GBM\", \"Default Model\", \"Ensemble\"]\n",
    "# def withinNPercent(df,n=20,default_column=\"Ridge\"):\n",
    "#     col=[val for val in df.columns if val not in columns]\n",
    "#     col.append(default_column)\n",
    "#     s1=df[col].loc[np.logical_and(df[default_column]<=n,df[default_column]>=-n),:]\n",
    "#     print(default_column,\"=\",s1.shape[0])\n",
    "#     print(round((s1.shape[0]/df.shape[0])*100,2),\"% of predictions by\",default_column,\"are within\",n,\"% of the true price.\")\n",
    "# for val in columns:\n",
    "#     withinNPercent(result,5,val)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hyperparameter Tuning "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "# performanceDataFrame=pd.DataFrame(columns=['Fit', 'Best Estimator','R2 score'])\n",
    "# #params is a dictionary with keys as the parameter and values as the list of values that we want to test for. \n",
    "# from sklearn.model_selection import GridSearchCV\n",
    "# def gridSearchCV(X, y, params, modelObject,performanceData):\n",
    "#     clf=GridSearchCV(modelObject,param_grid=params)\n",
    "#     clf.fit(X,y)\n",
    "#     performanceData=performanceData.append({'Fit':type(modelObject).__name__,'Best Estimator':clf.best_estimator_,'R2 score':clf.best_score_},ignore_index=True)\n",
    "#     return performanceData"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Ridge regression\n",
    "# X=strat1Array\n",
    "# y=y_log1\n",
    "# params={'alpha':[1,2,3]}\n",
    "# modelObject=Ridge()\n",
    "# performanceDataFrame=gridSearchCV(X,y,params,modelObject,performanceDataFrame)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Random Forest \n",
    "# X=strat2pca\n",
    "# y=y_log1\n",
    "# params={'n_estimators':[10,100,500,1000,2000],'max_depth':[10,100,500,1000],'max_features':['auto','sqrt','log2'],'bootstrap':['True','False']}\n",
    "# modelObject=RandomForestRegressor()\n",
    "# performanceDataFrame=gridSearchCV(X,y,params,modelObject,performanceDataFrame)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pd.DataFrame({\"Features\" : X.columns, \" Regression Coefficient\": lr.coef_ })"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [],
   "source": [
    "# def pipeLine():\n",
    "# # Inputs: Size of data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.preprocessing import PolynomialFeatures"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "# pf = PolynomialFeatures()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.neighbors import KNeighborsRegressor\n",
    "# knn = KNeighborsRegressor(n_neighbors=10)\n",
    "# knn.fit(X_train,y_train)\n"
   ]
  }
 ],
 "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": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}