Hackathon_FundRatings_RandomForest.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 229,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Importing all required libraries\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from sklearn.decomposition import PCA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 230,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/chai3/opt/anaconda3/lib/python3.7/site-packages/IPython/core/interactiveshell.py:3058: DtypeWarning: Columns (29,31,32,64,78,101) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  interactivity=interactivity, compiler=compiler, result=result)\n"
     ]
    }
   ],
   "source": [
    "#Reading the data file\n",
    "FundRating=pd.read_csv('Rating.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 231,
   "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>bb_rating</th>\n",
       "      <th>us_govt_bond_rating</th>\n",
       "      <th>below_b_rating</th>\n",
       "      <th>others_rating</th>\n",
       "      <th>maturity_bond</th>\n",
       "      <th>b_rating</th>\n",
       "      <th>a_rating</th>\n",
       "      <th>aaa_rating</th>\n",
       "      <th>aa_rating</th>\n",
       "      <th>bbb_rating</th>\n",
       "      <th>...</th>\n",
       "      <th>10_years_beta_fund</th>\n",
       "      <th>10yrs_treynor_ratio_fund</th>\n",
       "      <th>10_years_return_mean_annual_category</th>\n",
       "      <th>10yrs_treynor_ratio_category</th>\n",
       "      <th>10_years_return_fund</th>\n",
       "      <th>10_years_alpha_category</th>\n",
       "      <th>10_years_beta_category</th>\n",
       "      <th>10years_category_std</th>\n",
       "      <th>10_years_return_mean_annual_fund</th>\n",
       "      <th>10_years_return_category</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.21</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>14.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>1.08</td>\n",
       "      <td>15.57</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.15</td>\n",
       "      <td>17.25</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>1.42</td>\n",
       "      <td>15.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.95</td>\n",
       "      <td>16.58</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.15</td>\n",
       "      <td>16.21</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>1.33</td>\n",
       "      <td>15.94</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.84</td>\n",
       "      <td>16.38</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>14.12</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.13</td>\n",
       "      <td>1.16</td>\n",
       "      <td>13.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.12</td>\n",
       "      <td>NaN</td>\n",
       "      <td>11.53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>1.22</td>\n",
       "      <td>8.44</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.09</td>\n",
       "      <td>10.77</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.10</td>\n",
       "      <td>0.91</td>\n",
       "      <td>11.07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>17.94</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.78</td>\n",
       "      <td>10.04</td>\n",
       "      <td>NaN</td>\n",
       "      <td>21.07</td>\n",
       "      <td>8.53</td>\n",
       "      <td>7.44</td>\n",
       "      <td>1.64</td>\n",
       "      <td>32.55</td>\n",
       "      <td>...</td>\n",
       "      <td>0.49</td>\n",
       "      <td>7.08</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>3.90</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.07</td>\n",
       "      <td>0.33</td>\n",
       "      <td>4.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.14</td>\n",
       "      <td>0.00</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.13</td>\n",
       "      <td>NaN</td>\n",
       "      <td>13.68</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>...</td>\n",
       "      <td>0.91</td>\n",
       "      <td>-0.23</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.08</td>\n",
       "      <td>0.58</td>\n",
       "      <td>-0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.16</td>\n",
       "      <td>0.19</td>\n",
       "      <td>8.08</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>1.78</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.17</td>\n",
       "      <td>5.86</td>\n",
       "      <td>7.34</td>\n",
       "      <td>1.35</td>\n",
       "      <td>8.40</td>\n",
       "      <td>63.07</td>\n",
       "      <td>3.90</td>\n",
       "      <td>15.47</td>\n",
       "      <td>...</td>\n",
       "      <td>0.90</td>\n",
       "      <td>11.14</td>\n",
       "      <td>0.01</td>\n",
       "      <td>-0.14</td>\n",
       "      <td>10.53</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.05</td>\n",
       "      <td>0.87</td>\n",
       "      <td>6.24</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10 rows × 126 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   bb_rating  us_govt_bond_rating  below_b_rating  others_rating  \\\n",
       "0       0.00                  0.0            0.00           0.00   \n",
       "1       0.00                  0.0            0.00           0.00   \n",
       "2       0.00                  0.0            0.00           0.00   \n",
       "3       0.00                  0.0            0.00           0.00   \n",
       "4       0.00                  0.0            0.00           0.00   \n",
       "5       0.00                  0.0            0.00           0.00   \n",
       "6      17.94                  0.0            0.78          10.04   \n",
       "7       0.00                  0.0            0.00           0.00   \n",
       "8       0.00                  0.0            0.00           0.00   \n",
       "9       1.78                  0.0            0.17           5.86   \n",
       "\n",
       "   maturity_bond  b_rating  a_rating  aaa_rating  aa_rating  bbb_rating  ...  \\\n",
       "0            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "1            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "2            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "3            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "4            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "5            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "6            NaN     21.07      8.53        7.44       1.64       32.55  ...   \n",
       "7            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "8            NaN      0.00      0.00        0.00       0.00        0.00  ...   \n",
       "9           7.34      1.35      8.40       63.07       3.90       15.47  ...   \n",
       "\n",
       "   10_years_beta_fund  10yrs_treynor_ratio_fund  \\\n",
       "0                 NaN                       NaN   \n",
       "1                1.08                     15.57   \n",
       "2                0.95                     16.58   \n",
       "3                0.84                     16.38   \n",
       "4                 NaN                       NaN   \n",
       "5                1.22                      8.44   \n",
       "6                0.49                      7.08   \n",
       "7                 NaN                       NaN   \n",
       "8                0.91                     -0.23   \n",
       "9                0.90                     11.14   \n",
       "\n",
       "   10_years_return_mean_annual_category  10yrs_treynor_ratio_category  \\\n",
       "0                                  0.01                          0.21   \n",
       "1                                  0.01                          0.15   \n",
       "2                                  0.01                          0.15   \n",
       "3                                  0.01                          0.14   \n",
       "4                                  0.01                          0.08   \n",
       "5                                  0.01                          0.09   \n",
       "6                                  0.00                          0.06   \n",
       "7                                  0.01                          0.14   \n",
       "8                                  0.01                          0.08   \n",
       "9                                  0.01                         -0.14   \n",
       "\n",
       "   10_years_return_fund  10_years_alpha_category  10_years_beta_category  \\\n",
       "0                  0.00                     0.06                    0.01   \n",
       "1                 17.25                    -0.01                    0.01   \n",
       "2                 16.21                    -0.01                    0.01   \n",
       "3                 14.12                    -0.02                    0.01   \n",
       "4                  0.00                    -0.02                    0.01   \n",
       "5                 10.77                     0.00                    0.01   \n",
       "6                  3.90                    -0.01                    0.01   \n",
       "7                  0.00                    -0.02                    0.01   \n",
       "8                  0.58                    -0.01                    0.01   \n",
       "9                 10.53                     0.01                    0.00   \n",
       "\n",
       "   10years_category_std  10_years_return_mean_annual_fund  \\\n",
       "0                  0.13                               NaN   \n",
       "1                  0.14                              1.42   \n",
       "2                  0.14                              1.33   \n",
       "3                  0.13                              1.16   \n",
       "4                  0.12                               NaN   \n",
       "5                  0.10                              0.91   \n",
       "6                  0.07                              0.33   \n",
       "7                  0.13                               NaN   \n",
       "8                  0.16                              0.19   \n",
       "9                  0.05                              0.87   \n",
       "\n",
       "   10_years_return_category  \n",
       "0                     14.30  \n",
       "1                     15.94  \n",
       "2                     15.94  \n",
       "3                     13.68  \n",
       "4                     11.53  \n",
       "5                     11.07  \n",
       "6                      4.77  \n",
       "7                     13.68  \n",
       "8                      8.08  \n",
       "9                      6.24  \n",
       "\n",
       "[10 rows x 126 columns]"
      ]
     },
     "execution_count": 231,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FundRating.head(10)\n",
    "#From below data, we can see that most of the values are numeric. We do have NAN values present in the data set\n",
    "#Let's analyze it further"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 232,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Data pre-processing – Perform all the necessary pre-processing on the data ready to be fed to an Unsupervised algorithm "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 233,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 25000 entries, 0 to 24999\n",
      "Columns: 126 entries, bb_rating to 10_years_return_category\n",
      "dtypes: float64(109), int64(1), object(16)\n",
      "memory usage: 24.0+ MB\n"
     ]
    }
   ],
   "source": [
    "FundRating.info()\n",
    "#The data set has 846 rows and 19 columns, all are numeric except column 'class' which is of type object. \n",
    "#This indicates that there is categorical variable which needs to represented in numeric format.\n",
    "#From below data it is evident that not all 846 rows has values, some cells are empty "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 234,
   "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>count</th>\n",
       "      <th>mean</th>\n",
       "      <th>std</th>\n",
       "      <th>min</th>\n",
       "      <th>25%</th>\n",
       "      <th>50%</th>\n",
       "      <th>75%</th>\n",
       "      <th>max</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>bb_rating</td>\n",
       "      <td>24886.0</td>\n",
       "      <td>3.473804</td>\n",
       "      <td>8.299747</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>2.45</td>\n",
       "      <td>80.47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>us_govt_bond_rating</td>\n",
       "      <td>24886.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>below_b_rating</td>\n",
       "      <td>24886.0</td>\n",
       "      <td>0.827525</td>\n",
       "      <td>2.700019</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.10</td>\n",
       "      <td>42.30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>others_rating</td>\n",
       "      <td>24886.0</td>\n",
       "      <td>1.666769</td>\n",
       "      <td>6.885158</td>\n",
       "      <td>-68.21</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.33</td>\n",
       "      <td>100.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>maturity_bond</td>\n",
       "      <td>8093.0</td>\n",
       "      <td>7.765444</td>\n",
       "      <td>4.148558</td>\n",
       "      <td>0.00</td>\n",
       "      <td>5.46</td>\n",
       "      <td>7.29</td>\n",
       "      <td>8.92</td>\n",
       "      <td>29.30</td>\n",
       "    </tr>\n",
       "    <tr>\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",
       "      <td>10_years_alpha_category</td>\n",
       "      <td>24894.0</td>\n",
       "      <td>0.001136</td>\n",
       "      <td>0.027795</td>\n",
       "      <td>-0.11</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10_years_beta_category</td>\n",
       "      <td>24894.0</td>\n",
       "      <td>0.009076</td>\n",
       "      <td>0.005493</td>\n",
       "      <td>-0.02</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.01</td>\n",
       "      <td>0.12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10years_category_std</td>\n",
       "      <td>24894.0</td>\n",
       "      <td>0.109399</td>\n",
       "      <td>0.055202</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.06</td>\n",
       "      <td>0.12</td>\n",
       "      <td>0.15</td>\n",
       "      <td>0.34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10_years_return_mean_annual_fund</td>\n",
       "      <td>16416.0</td>\n",
       "      <td>0.858579</td>\n",
       "      <td>0.458993</td>\n",
       "      <td>-3.64</td>\n",
       "      <td>0.53</td>\n",
       "      <td>0.90</td>\n",
       "      <td>1.20</td>\n",
       "      <td>3.28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10_years_return_category</td>\n",
       "      <td>24885.0</td>\n",
       "      <td>9.679289</td>\n",
       "      <td>4.926417</td>\n",
       "      <td>-24.99</td>\n",
       "      <td>6.44</td>\n",
       "      <td>9.97</td>\n",
       "      <td>14.12</td>\n",
       "      <td>18.72</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>110 rows × 8 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                    count      mean       std    min   25%  \\\n",
       "bb_rating                         24886.0  3.473804  8.299747   0.00  0.00   \n",
       "us_govt_bond_rating               24886.0  0.000000  0.000000   0.00  0.00   \n",
       "below_b_rating                    24886.0  0.827525  2.700019  -0.02  0.00   \n",
       "others_rating                     24886.0  1.666769  6.885158 -68.21  0.00   \n",
       "maturity_bond                      8093.0  7.765444  4.148558   0.00  5.46   \n",
       "...                                   ...       ...       ...    ...   ...   \n",
       "10_years_alpha_category           24894.0  0.001136  0.027795  -0.11 -0.02   \n",
       "10_years_beta_category            24894.0  0.009076  0.005493  -0.02  0.01   \n",
       "10years_category_std              24894.0  0.109399  0.055202   0.00  0.06   \n",
       "10_years_return_mean_annual_fund  16416.0  0.858579  0.458993  -3.64  0.53   \n",
       "10_years_return_category          24885.0  9.679289  4.926417 -24.99  6.44   \n",
       "\n",
       "                                   50%    75%     max  \n",
       "bb_rating                         0.00   2.45   80.47  \n",
       "us_govt_bond_rating               0.00   0.00    0.00  \n",
       "below_b_rating                    0.00   0.10   42.30  \n",
       "others_rating                     0.00   0.33  100.00  \n",
       "maturity_bond                     7.29   8.92   29.30  \n",
       "...                                ...    ...     ...  \n",
       "10_years_alpha_category           0.00   0.01    0.10  \n",
       "10_years_beta_category            0.01   0.01    0.12  \n",
       "10years_category_std              0.12   0.15    0.34  \n",
       "10_years_return_mean_annual_fund  0.90   1.20    3.28  \n",
       "10_years_return_category          9.97  14.12   18.72  \n",
       "\n",
       "[110 rows x 8 columns]"
      ]
     },
     "execution_count": 234,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FundRating.describe().transpose()\n",
    "#Inference: From below data we can conclude that outlier values  and negetive values are present in data set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 235,
   "metadata": {},
   "outputs": [],
   "source": [
    "FundRating=FundRating.replace(\",\", \"\",regex=True)\n",
    "FundRating=FundRating.replace(\">\", \"\",regex=True)\n",
    "FundRating=FundRating.replace(\"<\",\"\",regex=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 236,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "FundRating[\"pb_ratio\"]=np.where(FundRating[\"pb_ratio\"]>(5.085), (5.085), FundRating[\"pb_ratio\"])\n",
    "FundRating[\"pb_ratio\"]=np.where(FundRating[\"pb_ratio\"]<(-2.155), (-2.155), FundRating[\"pb_ratio\"])\n",
    "\n",
    "FundRating[\"fund_ratio_net_annual_expense\"]=np.where(FundRating[\"fund_ratio_net_annual_expense\"]>(2.52), (2.52), FundRating[\"fund_ratio_net_annual_expense\"])\n",
    "FundRating[\"fund_ratio_net_annual_expense\"]=np.where(FundRating[\"fund_ratio_net_annual_expense\"]<(-0.3600), (-0.3600), FundRating[\"fund_ratio_net_annual_expense\"])\n",
    "\n",
    "FundRating[\"portfolio_cyclical_consumer_allocation\"]=np.where(FundRating[\"portfolio_cyclical_consumer_allocation\"]>(33.0), (33.0), FundRating[\"portfolio_cyclical_consumer_allocation\"])\n",
    "FundRating[\"portfolio_cyclical_consumer_allocation\"]=np.where(FundRating[\"portfolio_cyclical_consumer_allocation\"]<(-19.799), (-19.799), FundRating[\"portfolio_cyclical_consumer_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_utils_allocation\"]=np.where(FundRating[\"portfolio_utils_allocation\"]>(8.35), (8.35), FundRating[\"portfolio_utils_allocation\"])\n",
    "FundRating[\"portfolio_utils_allocation\"]=np.where(FundRating[\"portfolio_utils_allocation\"]<(-5.01), (-5.01), FundRating[\"portfolio_utils_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_property_allocation\"]=np.where(FundRating[\"portfolio_property_allocation\"]>(11.08125), (11.08125), FundRating[\"portfolio_property_allocation\"])\n",
    "FundRating[\"portfolio_property_allocation\"]=np.where(FundRating[\"portfolio_property_allocation\"]<(-6.6487), (-6.6487), FundRating[\"portfolio_property_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_healthcare_allocation\"]=np.where(FundRating[\"portfolio_healthcare_allocation\"]>(33.875), (33.875), FundRating[\"portfolio_healthcare_allocation\"])\n",
    "FundRating[\"portfolio_healthcare_allocation\"]=np.where(FundRating[\"portfolio_healthcare_allocation\"]<(-20.3250), (-20.3250), FundRating[\"portfolio_healthcare_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_consumer_defence_allocation\"]=np.where(FundRating[\"portfolio_consumer_defence_allocation\"]>(19.675), (19.675), FundRating[\"portfolio_consumer_defence_allocation\"])\n",
    "FundRating[\"portfolio_consumer_defence_allocation\"]=np.where(FundRating[\"portfolio_consumer_defence_allocation\"]<(-11.805), (-11.805), FundRating[\"portfolio_consumer_defence_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_energy_allocation\"]=np.where(FundRating[\"portfolio_energy_allocation\"]>(15.5562), (15.5562), FundRating[\"portfolio_energy_allocation\"])\n",
    "FundRating[\"portfolio_energy_allocation\"]=np.where(FundRating[\"portfolio_energy_allocation\"]<(-9.3337), (-9.3337), FundRating[\"portfolio_energy_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_materials_basic_allocation\"]=np.where(FundRating[\"portfolio_materials_basic_allocation\"]>(12.625), (12.625), FundRating[\"portfolio_materials_basic_allocation\"])\n",
    "FundRating[\"portfolio_materials_basic_allocation\"]=np.where(FundRating[\"portfolio_materials_basic_allocation\"]<(-7.5749), (-7.5749), FundRating[\"portfolio_materials_basic_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_tech_allocation\"]=np.where(FundRating[\"portfolio_tech_allocation\"]>(48.97499), (48.97499), FundRating[\"portfolio_tech_allocation\"])\n",
    "FundRating[\"portfolio_tech_allocation\"]=np.where(FundRating[\"portfolio_tech_allocation\"]<(-29.3849), (-29.3849), FundRating[\"portfolio_tech_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_industrials_allocation\"]=np.where(FundRating[\"portfolio_industrials_allocation\"]>(31.825), (31.825), FundRating[\"portfolio_industrials_allocation\"])\n",
    "FundRating[\"portfolio_industrials_allocation\"]=np.where(FundRating[\"portfolio_industrials_allocation\"]<(-19.095), (-19.095), FundRating[\"portfolio_industrials_allocation\"])\n",
    "\n",
    "FundRating[\"portfolio_financial_services\"]=np.where(FundRating[\"portfolio_financial_services\"]>(44.75), (44.75), FundRating[\"portfolio_financial_services\"])\n",
    "FundRating[\"portfolio_financial_services\"]=np.where(FundRating[\"portfolio_financial_services\"]<(-26.8499), (-26.8499), FundRating[\"portfolio_financial_services\"])\n",
    "\n",
    "FundRating[\"portfolio_communication_allocation\"]=np.where(FundRating[\"portfolio_communication_allocation\"]>(8.5), (8.5), FundRating[\"portfolio_communication_allocation\"])\n",
    "FundRating[\"portfolio_communication_allocation\"]=np.where(FundRating[\"portfolio_communication_allocation\"]<(-5.1), (-5.1), FundRating[\"portfolio_communication_allocation\"])\n",
    "    \n",
    "FundRating[\"duration_bond\"]=np.where(FundRating[\"duration_bond\"]>(4.8), (4.8), FundRating[\"duration_bond\"])\n",
    "FundRating[\"duration_bond\"]=np.where(FundRating[\"duration_bond\"]<(4.8), (4.8), FundRating[\"duration_bond\"])\n",
    "\n",
    "FundRating[\"bbb_rating\"]=np.where(FundRating[\"bbb_rating\"]>(28.3499), (28.3499), FundRating[\"bbb_rating\"])\n",
    "FundRating[\"bbb_rating\"]=np.where(FundRating[\"bbb_rating\"]<(-17.009), (-17.009), FundRating[\"bbb_rating\"])\n",
    "\n",
    "FundRating[\"aa_rating\"]=np.where(FundRating[\"aa_rating\"]>(7.4750), (7.4750), FundRating[\"aa_rating\"])\n",
    "FundRating[\"aa_rating\"]=np.where(FundRating[\"aa_rating\"]<(-4.485), (-4.485), FundRating[\"aa_rating\"])\n",
    "\n",
    "FundRating[\"aaa_rating\"]=np.where(FundRating[\"aaa_rating\"]>(46.0500), (46.0500), FundRating[\"aaa_rating\"])\n",
    "FundRating[\"aaa_rating\"]=np.where(FundRating[\"aaa_rating\"]<(-27.6300), (-27.6300), FundRating[\"aaa_rating\"])\n",
    "\n",
    "FundRating[\"a_rating\"]=np.where(FundRating[\"a_rating\"]>(20.6937), (20.6937), FundRating[\"a_rating\"])\n",
    "FundRating[\"a_rating\"]=np.where(FundRating[\"a_rating\"]<(-12.4162), (-12.4162), FundRating[\"a_rating\"])\n",
    "\n",
    "FundRating[\"maturity_bond\"]=np.where(FundRating[\"maturity_bond\"]>(7.29), (7.29), FundRating[\"maturity_bond\"])\n",
    "FundRating[\"maturity_bond\"]=np.where(FundRating[\"maturity_bond\"]<(7.29), (7.29), FundRating[\"maturity_bond\"])\n",
    "\n",
    "FundRating[\"others_rating\"]=np.where(FundRating[\"others_rating\"]>(0.8), (0.8), FundRating[\"others_rating\"])\n",
    "FundRating[\"others_rating\"]=np.where(FundRating[\"others_rating\"]<(-0.48), (-0.48), FundRating[\"others_rating\"])\n",
    "\n",
    "FundRating[\"2013_return_fund\"]=np.where(FundRating[\"2013_return_fund\"]>(47.7037), (47.7037), FundRating[\"2013_return_fund\"])\n",
    "FundRating[\"2013_return_fund\"]=np.where(FundRating[\"2013_return_fund\"]<(-13.1862), (-13.1862), FundRating[\"2013_return_fund\"])\n",
    "\n",
    "FundRating[\"fund_return_3months\"]=np.where(FundRating[\"fund_return_3months\"]>(26.2000), (26.2000), FundRating[\"fund_return_3months\"])\n",
    "FundRating[\"fund_return_3months\"]=np.where(FundRating[\"fund_return_3months\"]<(-9.0), (-9.0), FundRating[\"fund_return_3months\"])\n",
    "\n",
    "FundRating[\"ytd_return_category\"]=np.where(FundRating[\"ytd_return_category\"]>(24.82), (24.82), FundRating[\"ytd_return_category\"])\n",
    "FundRating[\"ytd_return_category\"]=np.where(FundRating[\"ytd_return_category\"]<(-6.8599), (-6.8599), FundRating[\"ytd_return_category\"])\n",
    "\n",
    "FundRating[\"2017_category_return\"]=np.where(FundRating[\"2017_category_return\"]>(44.045), (44.045), FundRating[\"2017_category_return\"])\n",
    "FundRating[\"2017_category_return\"]=np.where(FundRating[\"2017_category_return\"]<(-16.0750), (-16.0750), FundRating[\"2017_category_return\"])\n",
    "\n",
    "FundRating[\"1_year_return_fund\"]=np.where(FundRating[\"1_year_return_fund\"]>(12.795), (12.795), FundRating[\"1_year_return_fund\"])\n",
    "FundRating[\"1_year_return_fund\"]=np.where(FundRating[\"1_year_return_fund\"]<(-7.7250), (-7.7250), FundRating[\"1_year_return_fund\"])\n",
    "\n",
    "FundRating[\"2015_return_fund\"]=np.where(FundRating[\"2015_return_fund\"]>(4.79), (4.79), FundRating[\"2015_return_fund\"])\n",
    "FundRating[\"2015_return_fund\"]=np.where(FundRating[\"2015_return_fund\"]<(-7.21), (-7.21), FundRating[\"2015_return_fund\"])\n",
    "\n",
    "FundRating[\"portfolio_convertable\"]=np.where(FundRating[\"portfolio_convertable\"]>(0.1750), (0.1750), FundRating[\"portfolio_convertable\"])\n",
    "FundRating[\"portfolio_convertable\"]=np.where(FundRating[\"portfolio_convertable\"]<(-0.1050), (-0.1050), FundRating[\"portfolio_convertable\"])\n",
    "\n",
    "FundRating[\"3_months_return_category\"]=np.where(FundRating[\"3_months_return_category\"]>(24.82), (24.82), FundRating[\"3_months_return_category\"])\n",
    "FundRating[\"3_months_return_category\"]=np.where(FundRating[\"3_months_return_category\"]<(-6.8599), (-6.8599), FundRating[\"3_months_return_category\"])\n",
    "\n",
    "FundRating[\"portfolio_others\"]=np.where(FundRating[\"portfolio_others\"]>(0.7000), (0.7000), FundRating[\"portfolio_others\"])\n",
    "FundRating[\"portfolio_others\"]=np.where(FundRating[\"portfolio_others\"]<(-0.4200), (-0.4200), FundRating[\"portfolio_others\"])\n",
    "\n",
    "FundRating[\"2016_return_fund\"]=np.where(FundRating[\"2016_return_fund\"]>(19.7200), (19.7200), FundRating[\"2016_return_fund\"])\n",
    "FundRating[\"2016_return_fund\"]=np.where(FundRating[\"2016_return_fund\"]<(-7.160), (-7.160), FundRating[\"2016_return_fund\"])\n",
    "\n",
    "FundRating[\"stock_percent_of_portfolio\"]=np.where(FundRating[\"stock_percent_of_portfolio\"]>(243.11), (243.11), FundRating[\"stock_percent_of_portfolio\"])\n",
    "FundRating[\"stock_percent_of_portfolio\"]=np.where(FundRating[\"stock_percent_of_portfolio\"]<(-144.95), (-144.95), FundRating[\"stock_percent_of_portfolio\"])\n",
    "\n",
    "FundRating[\"2016_return_category\"]=np.where(FundRating[\"2016_return_category\"]>(18.28), (18.28), FundRating[\"2016_return_category\"])\n",
    "FundRating[\"2016_return_category\"]=np.where(FundRating[\"2016_return_category\"]<(-5.7999), (-5.7999), FundRating[\"2016_return_category\"])\n",
    "\n",
    "FundRating[\"2011_return_fund\"]=np.where(FundRating[\"2011_return_fund\"]>(2.945), (2.945), FundRating[\"2011_return_fund\"])\n",
    "FundRating[\"2011_return_fund\"]=np.where(FundRating[\"2011_return_fund\"]<(-4.055), (-4.055), FundRating[\"2011_return_fund\"])\n",
    "\n",
    "FundRating[\"2010_return_fund\"]=np.where(FundRating[\"2010_return_fund\"]>(13.5962), (13.5962), FundRating[\"2010_return_fund\"])\n",
    "FundRating[\"2010_return_fund\"]=np.where(FundRating[\"2010_return_fund\"]<(12.6), (12.6), FundRating[\"2010_return_fund\"])\n",
    "\n",
    "FundRating[\"fund_return_3years\"]=np.where(FundRating[\"fund_return_3years\"]>(21.2299), (21.2299), FundRating[\"fund_return_3years\"])\n",
    "FundRating[\"fund_return_3years\"]=np.where(FundRating[\"fund_return_3years\"]<(-8.2099), (-8.2099), FundRating[\"fund_return_3years\"])\n",
    "\n",
    "FundRating[\"2012_fund_return\"]=np.where(FundRating[\"2012_fund_return\"]>(18.4049), (18.4049), FundRating[\"2012_fund_return\"])\n",
    "FundRating[\"2012_fund_return\"]=np.where(FundRating[\"2012_fund_return\"]<(8.5250), (8.5250), FundRating[\"2012_fund_return\"])\n",
    "\n",
    "FundRating[\"2018_return_fund\"]=np.where(FundRating[\"2018_return_fund\"]>(10.73), (10.73), FundRating[\"2018_return_fund\"])\n",
    "FundRating[\"2018_return_fund\"]=np.where(FundRating[\"2018_return_fund\"]<(-22.79), (-22.79), FundRating[\"2018_return_fund\"])\n",
    "\n",
    "FundRating[\"2017_return_fund\"]=np.where(FundRating[\"2017_return_fund\"]>(42.2237), (42.2237), FundRating[\"2017_return_fund\"])\n",
    "FundRating[\"2017_return_fund\"]=np.where(FundRating[\"2017_return_fund\"]<(-14.1062), (-14.1062), FundRating[\"2017_return_fund\"])\n",
    "\n",
    "FundRating[\"category_ratio_net_annual_expense\"]=np.where(FundRating[\"category_ratio_net_annual_expense\"]>(1.7349), (1.7349), FundRating[\"category_ratio_net_annual_expense\"])\n",
    "FundRating[\"category_ratio_net_annual_expense\"]=np.where(FundRating[\"category_ratio_net_annual_expense\"]<(0.2550), (0.2550), FundRating[\"category_ratio_net_annual_expense\"])\n",
    "\n",
    "FundRating[\"category_return_2015\"]=np.where(FundRating[\"category_return_2015\"]>(1.4950), (1.4950), FundRating[\"category_return_2015\"])\n",
    "FundRating[\"category_return_2015\"]=np.where(FundRating[\"category_return_2015\"]<(-5.025), (-5.025), FundRating[\"category_return_2015\"])\n",
    "\n",
    "FundRating[\"1_month_fund_return\"]=np.where(FundRating[\"1_month_fund_return\"]>(3.75), (3.75), FundRating[\"1_month_fund_return\"])\n",
    "FundRating[\"1_month_fund_return\"]=np.where(FundRating[\"1_month_fund_return\"]<(-1.69), (-1.69), FundRating[\"1_month_fund_return\"])\n",
    "\n",
    "FundRating[\"bond_percentage_of_porfolio\"]=np.where(FundRating[\"bond_percentage_of_porfolio\"]>(159.25), (159.25), FundRating[\"bond_percentage_of_porfolio\"])\n",
    "FundRating[\"bond_percentage_of_porfolio\"]=np.where(FundRating[\"bond_percentage_of_porfolio\"]<(-95.5500), (-95.5500), FundRating[\"bond_percentage_of_porfolio\"])\n",
    "\n",
    "FundRating[\"portfolio_preferred\"]=np.where(FundRating[\"portfolio_preferred\"]>(0.025), (0.025), FundRating[\"portfolio_preferred\"])\n",
    "FundRating[\"portfolio_preferred\"]=np.where(FundRating[\"portfolio_preferred\"]<(-0.015), (-0.015), FundRating[\"portfolio_preferred\"])\n",
    "\n",
    "FundRating[\"2010_return_category\"]=np.where(FundRating[\"2010_return_category\"]>(14.115), (14.115), FundRating[\"2010_return_category\"])\n",
    "FundRating[\"2010_return_category\"]=np.where(FundRating[\"2010_return_category\"]<(13.115), (13.115), FundRating[\"2010_return_category\"])\n",
    "\n",
    "FundRating[\"2013_category_return\"]=np.where(FundRating[\"2013_category_return\"]>(52.13), (52.13), FundRating[\"2013_category_return\"])\n",
    "FundRating[\"2013_category_return\"]=np.where(FundRating[\"2013_category_return\"]<(-19.71), (-19.71), FundRating[\"2013_category_return\"])\n",
    "\n",
    "FundRating[\"3_years_alpha_fund\"]=np.where(FundRating[\"3_years_alpha_fund\"]>(4.6599), (4.6599), FundRating[\"3_years_alpha_fund\"])\n",
    "FundRating[\"3_years_alpha_fund\"]=np.where(FundRating[\"3_years_alpha_fund\"]<(-5.8999), (-5.8999), FundRating[\"3_years_alpha_fund\"])\n",
    "\n",
    "FundRating[\"3years_category_std\"]=np.where(FundRating[\"3years_category_std\"]>(0.265), (0.265), FundRating[\"3years_category_std\"])\n",
    "FundRating[\"3years_category_std\"]=np.where(FundRating[\"3years_category_std\"]<(-0.095), (-0.095), FundRating[\"3years_category_std\"])\n",
    "\n",
    "FundRating[\"3yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"3yrs_sharpe_ratio_fund\"]>(1.6849), (1.6849), FundRating[\"3yrs_sharpe_ratio_fund\"])\n",
    "FundRating[\"3yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"3yrs_sharpe_ratio_fund\"]<(-0.2749), (-0.2749), FundRating[\"3yrs_sharpe_ratio_fund\"])\n",
    "\n",
    "FundRating[\"3yrs_treynor_ratio_category\"]=np.where(FundRating[\"3yrs_treynor_ratio_category\"]>(0.1750), (0.1750), FundRating[\"3yrs_treynor_ratio_category\"])\n",
    "FundRating[\"3yrs_treynor_ratio_category\"]=np.where(FundRating[\"3yrs_treynor_ratio_category\"]<(-0.0250), (-0.0250), FundRating[\"3yrs_treynor_ratio_category\"])\n",
    "\n",
    "FundRating[\"3_years_return_mean_annual_fund\"]=np.where(FundRating[\"3_years_return_mean_annual_fund\"]>(1.635), (1.635), FundRating[\"3_years_return_mean_annual_fund\"])\n",
    "FundRating[\"3_years_return_mean_annual_fund\"]=np.where(FundRating[\"3_years_return_mean_annual_fund\"]<(-0.405), (-0.405), FundRating[\"3_years_return_mean_annual_fund\"])\n",
    "\n",
    "FundRating[\"fund_beta_3years\"]=np.where(FundRating[\"fund_beta_3years\"]>(1.6150), (1.6150), FundRating[\"fund_beta_3years\"])\n",
    "FundRating[\"fund_beta_3years\"]=np.where(FundRating[\"fund_beta_3years\"]<(0.3), (0.3), FundRating[\"fund_beta_3years\"])\n",
    "\n",
    "FundRating[\"3years_fund_r_squared\"]=np.where(FundRating[\"3years_fund_r_squared\"]>(131.43), (131.43), FundRating[\"3years_fund_r_squared\"])\n",
    "FundRating[\"3years_fund_r_squared\"]=np.where(FundRating[\"3years_fund_r_squared\"]<(26.55), (26.55), FundRating[\"3years_fund_r_squared\"])\n",
    "\n",
    "FundRating[\"3years_fund_std\"]=np.where(FundRating[\"3years_fund_std\"]>(23.68), (23.68), FundRating[\"3years_fund_std\"])\n",
    "FundRating[\"3years_fund_std\"]=np.where(FundRating[\"3years_fund_std\"]<(-6.9599), (-6.9599), FundRating[\"3years_fund_std\"])\n",
    "\n",
    "FundRating[\"category_beta_3years\"]=np.where(FundRating[\"category_beta_3years\"]>(0.01), (0.01), FundRating[\"category_beta_3years\"])\n",
    "FundRating[\"category_beta_3years\"]=np.where(FundRating[\"category_beta_3years\"]<(0.01), (0.01), FundRating[\"category_beta_3years\"])\n",
    "\n",
    "FundRating[\"fund_return_3years\"]=np.where(FundRating[\"fund_return_3years\"]>(21.2299), (21.2299), FundRating[\"fund_return_3years\"])\n",
    "FundRating[\"fund_return_3years\"]=np.where(FundRating[\"fund_return_3years\"]<(-8.2099), (-8.2099), FundRating[\"fund_return_3years\"])\n",
    "\n",
    "FundRating[\"3_years_alpha_category\"]=np.where(FundRating[\"3_years_alpha_category\"]>(0.015), (0.015), FundRating[\"3_years_alpha_category\"])\n",
    "FundRating[\"3_years_alpha_category\"]=np.where(FundRating[\"3_years_alpha_category\"]<(-0.025), (-0.025), FundRating[\"3_years_alpha_category\"])\n",
    "\n",
    "FundRating[\"3_years_return_mean_annual_category\"]=np.where(FundRating[\"3_years_return_mean_annual_category\"]>(0.025), (0.025), FundRating[\"3_years_return_mean_annual_category\"])\n",
    "FundRating[\"3_years_return_mean_annual_category\"]=np.where(FundRating[\"3_years_return_mean_annual_category\"]<(-0.015), (-0.015), FundRating[\"3_years_return_mean_annual_category\"])\n",
    "\n",
    "FundRating[\"3yrs_sharpe_ratio_category\"]=np.where(FundRating[\"3yrs_sharpe_ratio_category\"]>(0.025), (0.025), FundRating[\"3yrs_sharpe_ratio_category\"])\n",
    "FundRating[\"3yrs_sharpe_ratio_category\"]=np.where(FundRating[\"3yrs_sharpe_ratio_category\"]<(-0.015), (-0.015), FundRating[\"3yrs_sharpe_ratio_category\"])\n",
    "\n",
    "FundRating[\"3years_category_r_squared\"]=np.where(FundRating[\"3years_category_r_squared\"]>(1.28), (1.28), FundRating[\"3years_category_r_squared\"])\n",
    "FundRating[\"3years_category_r_squared\"]=np.where(FundRating[\"3years_category_r_squared\"]<(0.24), (0.24), FundRating[\"3years_category_r_squared\"])\n",
    "\n",
    "FundRating[\"3_years_return_category\"]=np.where(FundRating[\"3_years_return_category\"]>(18.485), (18.485), FundRating[\"3_years_return_category\"])\n",
    "FundRating[\"3_years_return_category\"]=np.where(FundRating[\"3_years_return_category\"]<(-4.1149), (-4.1149), FundRating[\"3_years_return_category\"])\n",
    "\n",
    "FundRating[\"category_r_squared_5years\"]=np.where(FundRating[\"category_r_squared_5years\"]>(1.2650), (1.2650), FundRating[\"category_r_squared_5years\"])\n",
    "FundRating[\"category_r_squared_5years\"]=np.where(FundRating[\"category_r_squared_5years\"]<(0.265), (0.265), FundRating[\"category_r_squared_5years\"])\n",
    "\n",
    "FundRating[\"5yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"5yrs_sharpe_ratio_fund\"]>(1.1549), (1.1549), FundRating[\"5yrs_sharpe_ratio_fund\"])\n",
    "FundRating[\"5yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"5yrs_sharpe_ratio_fund\"]<(-0.08499), (-0.08499), FundRating[\"5yrs_sharpe_ratio_fund\"])\n",
    "\n",
    "FundRating[\"5years_fund_r_squared\"]=np.where(FundRating[\"5years_fund_r_squared\"]>(125.6749), (125.6749), FundRating[\"5years_fund_r_squared\"])\n",
    "FundRating[\"5years_fund_r_squared\"]=np.where(FundRating[\"5years_fund_r_squared\"]<(33.475), (33.475), FundRating[\"5years_fund_r_squared\"])\n",
    "\n",
    "FundRating[\"5years_fund_std\"]=np.where(FundRating[\"5years_fund_std\"]>(22.4837), (22.4837), FundRating[\"5years_fund_std\"])\n",
    "FundRating[\"5years_fund_std\"]=np.where(FundRating[\"5years_fund_std\"]<(-4.5662), (-4.5662), FundRating[\"5years_fund_std\"])\n",
    "\n",
    "FundRating[\"5yrs_sharpe_ratio_category\"]=np.where(FundRating[\"5yrs_sharpe_ratio_category\"]>(0.025), (0.025), FundRating[\"5yrs_sharpe_ratio_category\"])\n",
    "FundRating[\"5yrs_sharpe_ratio_category\"]=np.where(FundRating[\"5yrs_sharpe_ratio_category\"]<(-0.015), (-0.015), FundRating[\"5yrs_sharpe_ratio_category\"])\n",
    "\n",
    "FundRating[\"5_years_alpha_fund\"]=np.where(FundRating[\"5_years_alpha_fund\"]>(3.560), (3.560), FundRating[\"5_years_alpha_fund\"])\n",
    "FundRating[\"5_years_alpha_fund\"]=np.where(FundRating[\"5_years_alpha_fund\"]<(-4.84), (-4.84), FundRating[\"5_years_alpha_fund\"])\n",
    "\n",
    "FundRating[\"5_years_beta_fund\"]=np.where(FundRating[\"5_years_beta_fund\"]>(1.4150), (1.4150), FundRating[\"5_years_beta_fund\"])\n",
    "FundRating[\"5_years_beta_fund\"]=np.where(FundRating[\"5_years_beta_fund\"]<(0.5), (0.5), FundRating[\"5_years_beta_fund\"])\n",
    "\n",
    "FundRating[\"5_years_return_mean_annual_fund\"]=np.where(FundRating[\"5_years_return_mean_annual_fund\"]>(1.0599), (1.0599), FundRating[\"5_years_return_mean_annual_fund\"])\n",
    "FundRating[\"5_years_return_mean_annual_fund\"]=np.where(FundRating[\"5_years_return_mean_annual_fund\"]<(-0.2199), (-0.2199), FundRating[\"5_years_return_mean_annual_fund\"])\n",
    "\n",
    "FundRating[\"5_years_return_mean_annual_category\"]=np.where(FundRating[\"5_years_return_mean_annual_category\"]>(0.025), (0.025), FundRating[\"5_years_return_mean_annual_category\"])\n",
    "FundRating[\"5_years_return_mean_annual_category\"]=np.where(FundRating[\"5_years_return_mean_annual_category\"]<(-0.015), (-0.015), FundRating[\"5_years_return_mean_annual_category\"])\n",
    "\n",
    "FundRating[\"5yrs_treynor_ratio_category\"]=np.where(FundRating[\"5yrs_treynor_ratio_category\"]>(0.12), (0.12), FundRating[\"5yrs_treynor_ratio_category\"])\n",
    "FundRating[\"5yrs_treynor_ratio_category\"]=np.where(FundRating[\"5yrs_treynor_ratio_category\"]<(-0.0399), (-0.0399), FundRating[\"5yrs_treynor_ratio_category\"])\n",
    "\n",
    "FundRating[\"5_years_return_fund\"]=np.where(FundRating[\"5_years_return_fund\"]>(14.375), (14.375), FundRating[\"5_years_return_fund\"])\n",
    "FundRating[\"5_years_return_fund\"]=np.where(FundRating[\"5_years_return_fund\"]<(-6.625), (-6.625), FundRating[\"5_years_return_fund\"])\n",
    "\n",
    "FundRating[\"5_years_alpha_category\"]=np.where(FundRating[\"5_years_alpha_category\"]>(0.03), (0.03), FundRating[\"5_years_alpha_category\"])\n",
    "FundRating[\"5_years_alpha_category\"]=np.where(FundRating[\"5_years_alpha_category\"]<(-0.05), (-0.05), FundRating[\"5_years_alpha_category\"])\n",
    "\n",
    "FundRating[\"5_years_beta_category\"]=np.where(FundRating[\"5_years_beta_category\"]>(0.01), (0.01), FundRating[\"5_years_beta_category\"])\n",
    "FundRating[\"5_years_beta_category\"]=np.where(FundRating[\"5_years_beta_category\"]<(0.01), (0.01), FundRating[\"5_years_beta_category\"])\n",
    "\n",
    "FundRating[\"5years_category_std\"]=np.where(FundRating[\"5years_category_std\"]>(0.25), (0.25), FundRating[\"5years_category_std\"])\n",
    "FundRating[\"5years_category_std\"]=np.where(FundRating[\"5years_category_std\"]<(-0.0699), (-0.0699), FundRating[\"5years_category_std\"])\n",
    "\n",
    "FundRating[\"10years_fund_r_squared\"]=np.where(FundRating[\"10years_fund_r_squared\"]>(105.2350), (105.2350), FundRating[\"10years_fund_r_squared\"])\n",
    "FundRating[\"10years_fund_r_squared\"]=np.where(FundRating[\"10years_fund_r_squared\"]<(66), (66), FundRating[\"10years_fund_r_squared\"])\n",
    "\n",
    "FundRating[\"10yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"10yrs_sharpe_ratio_fund\"]>(1.225), (1.225), FundRating[\"10yrs_sharpe_ratio_fund\"])\n",
    "FundRating[\"10yrs_sharpe_ratio_fund\"]=np.where(FundRating[\"10yrs_sharpe_ratio_fund\"]<(0.7050), (0.7050), FundRating[\"10yrs_sharpe_ratio_fund\"])\n",
    "\n",
    "FundRating[\"10_years_alpha_fund\"]=np.where(FundRating[\"10_years_alpha_fund\"]>(2.085), (2.085), FundRating[\"10_years_alpha_fund\"])\n",
    "FundRating[\"10_years_alpha_fund\"]=np.where(FundRating[\"10_years_alpha_fund\"]<(-2.675), (-2.675), FundRating[\"10_years_alpha_fund\"])\n",
    "\n",
    "FundRating[\"10years_fund_std\"]=np.where(FundRating[\"10years_fund_std\"]>(20.6299), (20.6299), FundRating[\"10years_fund_std\"])\n",
    "FundRating[\"10years_fund_std\"]=np.where(FundRating[\"10years_fund_std\"]<(3.2), (3.2), FundRating[\"10years_fund_std\"])\n",
    "\n",
    "FundRating[\"10yrs_sharpe_ratio_category\"]=np.where(FundRating[\"10yrs_sharpe_ratio_category\"]>(0.01), (0.01), FundRating[\"10yrs_sharpe_ratio_category\"])\n",
    "FundRating[\"10yrs_sharpe_ratio_category\"]=np.where(FundRating[\"10yrs_sharpe_ratio_category\"]<(0.01), (0.01), FundRating[\"10yrs_sharpe_ratio_category\"])\n",
    "\n",
    "FundRating[\"10_years_beta_fund\"]=np.where(FundRating[\"10_years_beta_fund\"]>(1.210), (1.210), FundRating[\"10_years_beta_fund\"])\n",
    "FundRating[\"10_years_beta_fund\"]=np.where(FundRating[\"10_years_beta_fund\"]<(0.8099), (0.8099), FundRating[\"10_years_beta_fund\"])\n",
    "\n",
    "FundRating[\"10_years_return_mean_annual_category\"]=np.where(FundRating[\"10_years_return_mean_annual_category\"]>(0.01), (0.01), FundRating[\"10_years_return_mean_annual_category\"])\n",
    "FundRating[\"10_years_return_mean_annual_category\"]=np.where(FundRating[\"10_years_return_mean_annual_category\"]<(0.01), (0.01), FundRating[\"10_years_return_mean_annual_category\"])\n",
    "\n",
    "FundRating[\"10yrs_treynor_ratio_category\"]=np.where(FundRating[\"10yrs_treynor_ratio_category\"]>(0.2450), (0.2450), FundRating[\"10yrs_treynor_ratio_category\"])\n",
    "FundRating[\"10yrs_treynor_ratio_category\"]=np.where(FundRating[\"10yrs_treynor_ratio_category\"]<(-0.035), (-0.035), FundRating[\"10yrs_treynor_ratio_category\"])\n",
    "\n",
    "FundRating[\"10_years_return_fund\"]=np.where(FundRating[\"10_years_return_fund\"]>(30.9), (30.9), FundRating[\"10_years_return_fund\"])\n",
    "FundRating[\"10_years_return_fund\"]=np.where(FundRating[\"10_years_return_fund\"]<(-18.54), (-18.54), FundRating[\"10_years_return_fund\"])\n",
    "\n",
    "FundRating[\"10_years_alpha_category\"]=np.where(FundRating[\"10_years_alpha_category\"]>(0.055), (0.055), FundRating[\"10_years_alpha_category\"])\n",
    "FundRating[\"10_years_alpha_category\"]=np.where(FundRating[\"10_years_alpha_category\"]<(-0.065), (-0.065), FundRating[\"10_years_alpha_category\"])\n",
    "\n",
    "\n",
    "FundRating[\"10_years_beta_category\"]=np.where(FundRating[\"10_years_beta_category\"]>(0.01), (0.01), FundRating[\"10_years_beta_category\"])\n",
    "FundRating[\"10_years_beta_category\"]=np.where(FundRating[\"10_years_beta_category\"]<(0.01), (0.01), FundRating[\"10_years_beta_category\"])\n",
    "\n",
    "FundRating[\"10years_category_std\"]=np.where(FundRating[\"10years_category_std\"]>(0.2850), (0.2850), FundRating[\"10years_category_std\"])\n",
    "FundRating[\"10years_category_std\"]=np.where(FundRating[\"10years_category_std\"]<(-0.0750), (-0.0750), FundRating[\"10years_category_std\"])\n",
    "\n",
    "FundRating[\"10_years_return_mean_annual_fund\"]=np.where(FundRating[\"10_years_return_mean_annual_fund\"]>(1.5250), (1.5250), FundRating[\"10_years_return_mean_annual_fund\"])\n",
    "FundRating[\"10_years_return_mean_annual_fund\"]=np.where(FundRating[\"10_years_return_mean_annual_fund\"]<(0.29), (0.29), FundRating[\"10_years_return_mean_annual_fund\"])\n",
    "\n",
    "FundRating[\"10_years_return_category\"]=np.where(FundRating[\"10_years_return_category\"]>(25.6399), (25.6399), FundRating[\"10_years_return_category\"])\n",
    "FundRating[\"10_years_return_category\"]=np.where(FundRating[\"10_years_return_category\"]<(-5.0799), (-5.0799), FundRating[\"10_years_return_category\"])\n",
    "\n",
    "FundRating[\"fund_return_3years.1\"]=np.where(FundRating[\"fund_return_3years.1\"]>(21.2299), (21.2299), FundRating[\"fund_return_3years.1\"])\n",
    "FundRating[\"fund_return_3years.1\"]=np.where(FundRating[\"fund_return_3years.1\"]<(-8.2099), (-8.2099), FundRating[\"fund_return_3years.1\"])\n",
    "\n",
    "FundRating[\"10years_category_r_squared\"]=np.where(FundRating[\"10years_category_r_squared\"]>(1.235), (1.235), FundRating[\"10years_category_r_squared\"])\n",
    "FundRating[\"10years_category_r_squared\"]=np.where(FundRating[\"10years_category_r_squared\"]<(0.4), (0.4), FundRating[\"10years_category_r_squared\"])\n",
    "\n",
    "FundRating[\"category_return_1month\"]=np.where(FundRating[\"category_return_1month\"]>(2.76), (2.76), FundRating[\"category_return_1month\"])\n",
    "FundRating[\"category_return_1month\"]=np.where(FundRating[\"category_return_1month\"]<(-0.9199), (-0.9199), FundRating[\"category_return_1month\"])\n",
    "\n",
    "FundRating[\"2014_return_fund\"]=np.where(FundRating[\"2014_return_fund\"]>(12.645), (12.645), FundRating[\"2014_return_fund\"])\n",
    "FundRating[\"2014_return_fund\"]=np.where(FundRating[\"2014_return_fund\"]<(-2.275), (-2.275), FundRating[\"2014_return_fund\"])\n",
    "\n",
    "FundRating[\"years_down\"]=np.where(FundRating[\"years_down\"]>(7.0), (7.0), FundRating[\"years_down\"])\n",
    "FundRating[\"years_down\"]=np.where(FundRating[\"years_down\"]<(-1.0), (-1.0), FundRating[\"years_down\"])\n",
    "\n",
    "\n",
    "FundRating[\"ytd_return_fund\"]=np.where(FundRating[\"ytd_return_fund\"]>(25.93), (25.93), FundRating[\"ytd_return_fund\"])\n",
    "FundRating[\"ytd_return_fund\"]=np.where(FundRating[\"ytd_return_fund\"]<(-8.39), (-8.39), FundRating[\"ytd_return_fund\"])\n",
    "\n",
    "FundRating[\"2011_return_category\"]=np.where(FundRating[\"2011_return_category\"]>(1.815), (1.815), FundRating[\"2011_return_category\"])\n",
    "FundRating[\"2011_return_category\"]=np.where(FundRating[\"2011_return_category\"]<(-5.025), (-5.025), FundRating[\"2011_return_category\"])\n",
    "\n",
    "FundRating[\"cash_percent_of_portfolio\"]=np.where(FundRating[\"cash_percent_of_portfolio\"]>(15.715), (15.715), FundRating[\"cash_percent_of_portfolio\"])\n",
    "FundRating[\"cash_percent_of_portfolio\"]=np.where(FundRating[\"cash_percent_of_portfolio\"]<(-7.445), (-7.445), FundRating[\"cash_percent_of_portfolio\"])\n",
    "\n",
    "FundRating[\"category_return_1year\"]=np.where(FundRating[\"category_return_1year\"]>(10.3099), (10.3099),FundRating[\"category_return_1year\"])\n",
    "FundRating[\"category_return_1year\"]=np.where(FundRating[\"category_return_1year\"]<(-5.1299), (-5.1299),FundRating[\"category_return_1year\"])\n",
    "\n",
    "FundRating[\"2018_return_category\"]=np.where(FundRating[\"2018_return_category\"]>(8.0950),(8.0950), FundRating[\"2018_return_category\"])\n",
    "FundRating[\"2018_return_category\"]=np.where(FundRating[\"2018_return_category\"]<(-19.0650),(-19.0650),FundRating[\"2018_return_category\"])\n",
    "\n",
    "FundRating[\"years_up\"]=np.where(FundRating[\"years_up\"]>(25.5), (25.5), FundRating[\"years_up\"])\n",
    "FundRating[\"years_up\"]=np.where(FundRating[\"years_up\"]<(-10.5), (-10.5), FundRating[\"years_up\"])\n",
    "\n",
    "\n",
    "FundRating[\"2012_return_category\"]=np.where(FundRating[\"2012_return_category\"]>(17.8550), (17.8550), FundRating[\"2012_return_category\"])\n",
    "FundRating[\"2012_return_category\"]=np.where(FundRating[\"2012_return_category\"]<(10.13), (10.14), FundRating[\"2012_return_category\"])\n",
    "\n",
    "FundRating[\"bb_rating\"]=np.where(FundRating[\"bb_rating\"]>(2.42), (6.05), FundRating[\"bb_rating\"])\n",
    "FundRating[\"bb_rating\"]=np.where(FundRating[\"bb_rating\"]<(-3.63), (-3.63), FundRating[\"bb_rating\"])\n",
    "\n",
    "FundRating[\"below_b_rating\"]=np.where(FundRating[\"below_b_rating\"]>(0.25),(0.25),FundRating[\"below_b_rating\"])\n",
    "FundRating[\"below_b_rating\"]=np.where(FundRating[\"below_b_rating\"]<(-0.1500), (-0.1500), FundRating[\"below_b_rating\"])\n",
    "\n",
    "FundRating[\"2014_category_return\"]=np.where(FundRating[\"2014_category_return\"]>(12.7399),(12.7399),FundRating[\"2014_category_return\"])\n",
    "FundRating[\"2014_category_return\"]=np.where(FundRating[\"2014_category_return\"]< (-3.1799), (-3.1799), FundRating[\"2014_category_return\"])\n",
    "\n",
    "FundRating[\"total_assets\"]=np.where(FundRating[\"total_assets\"]>(3908895000.0),(3908895000.0),FundRating[\"total_assets\"])\n",
    "FundRating[\"total_assets\"]=np.where(FundRating[\"total_assets\"]<(-2194825000.0), (-2194825000.0), FundRating[\"total_assets\"])\n",
    "\n",
    "FundRating[\"return_ytd\"]=np.where(FundRating[\"return_ytd\"]>(25.93),(25.93),FundRating['return_ytd'])\n",
    "FundRating[\"return_ytd\"]=np.where(FundRating[\"return_ytd\"]<(-8.39),(-8.39), FundRating['return_ytd'])\n",
    "\n",
    "FundRating[\"yield\"]=np.where(FundRating[\"yield\"]>(5.8800),(5.8800),FundRating['yield'])\n",
    "FundRating[\"yield\"]=np.where(FundRating[\"yield\"]<(-2.7600),(-2.7600), FundRating['yield'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 237,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(True, bb_rating                             114\n",
       " us_govt_bond_rating                   114\n",
       " below_b_rating                        114\n",
       " others_rating                         114\n",
       " maturity_bond                       16907\n",
       "                                     ...  \n",
       " 10_years_alpha_category               106\n",
       " 10_years_beta_category                106\n",
       " 10years_category_std                  106\n",
       " 10_years_return_mean_annual_fund     8584\n",
       " 10_years_return_category              115\n",
       " Length: 126, dtype: int64)"
      ]
     },
     "execution_count": 237,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Checking the presence of missing values.\n",
    "FundRating.isnull().values.any(),FundRating.isnull().sum()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 238,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a1f00c910>"
      ]
     },
     "execution_count": 238,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "#Let's reconfirm the understanding with heatmap\n",
    "sns.heatmap(FundRating.isnull(),yticklabels=False,cbar=False,cmap='viridis')\n",
    "#From below graph, it is evident that values are missing (indicated by yellow bar)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 239,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Labelling the variable based on size of the vehicle\n",
    "cleanup_nums = { \"greatstone_rating\": {0:\"zero\", 1:\"one\" , 2:\"two\", 3:\"three\", 4: \"four\", 5:\"five\" }\n",
    "                \n",
    "               }\n",
    "#Replacing in the data set\n",
    "FundRating.replace(cleanup_nums, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "\n",
    "from sklearn.base import TransformerMixin\n",
    "\n",
    "class DataFrameImputer(TransformerMixin):\n",
    "\n",
    "    def __init__(self):\n",
    "        \"\"\"Impute missing values.\n",
    "\n",
    "        Columns of dtype object are imputed with the most frequent value \n",
    "        in column.\n",
    "\n",
    "        Columns of other types are imputed with mean of column.\n",
    "\n",
    "        \"\"\"\n",
    "    def fit(self, FundRating, y=None):\n",
    "\n",
    "        self.fill = pd.Series([FundRating[c].value_counts().index[0]\n",
    "            if FundRating[c].dtype == np.dtype('O') else FundRating[c].median() for c in FundRating],\n",
    "            index=FundRating.columns)\n",
    "\n",
    "        return self\n",
    "\n",
    "    def transform(self, FundRating, y=None):\n",
    "        return FundRating.fillna(self.fill)\n",
    "    \n",
    "    \n",
    "    #X = pd.DataFrame(data)\n",
    "#xt = DataFrameImputer().fit_transform(FundRating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "FundRating = DataFrameImputer().fit_transform(FundRating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(False, False, Index([], dtype='object'))"
      ]
     },
     "execution_count": 242,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "FundRating.isnull().values.any(), FundRating.isna().values.any(), FundRating.columns[FundRating.isnull().any()]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1a20730ad0>"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "#Let's reconfirm the understanding with heatmap\n",
    "sns.heatmap(FundRating.isnull(),yticklabels=False,cbar=False,cmap='viridis')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 25000 entries, 0 to 24999\n",
      "Columns: 108 entries, bb_rating to 10_years_return_category\n",
      "dtypes: float64(107), int64(1)\n",
      "memory usage: 20.6 MB\n"
     ]
    }
   ],
   "source": [
    "FundRating1=FundRating\n",
    "FundRating=FundRating.drop(['greatstone_rating','fund_size','investment_class','mmc','pc_ratio','pe_ratio','ps_ratio.1','return_ytd','10yrs_treynor_ratio_fund','5yrs_treynor_ratio_fund','3yrs_treynor_ratio_fund','ps_ratio','inception_date','fund_id','category','fund_name','currency','parent_company'],axis=1)\n",
    "#FundRating=FundRating.replace(\",\", \"\")\n",
    "FundRating.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Labelling the variable based on size of the vehicle\n",
    "cleanup_nums = { \"investment_class\": {\"Blend\":1, \"Growth\":2 ,\"Value\":3 },\n",
    "                \n",
    "                 \"fund_size\":  {\"Small\": 1, \"Medium\": 2 ,\"Large\":3 }\n",
    "                \n",
    "                #\"greatstone_rating\": {\"zero\":0, \"one\":1 , \"two\":2, \"three\":3, \"four\":4, \"five\":5 }\n",
    "               }\n",
    "#Replacing in the data set\n",
    "FundRating.replace(cleanup_nums, inplace=True)\n",
    "\n",
    "#FundRating['class'].unique()\n",
    "#Inference: Labelling and replacement is reflecting successfully"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bb_rating\n",
      "us_govt_bond_rating\n",
      "below_b_rating\n",
      "others_rating\n",
      "maturity_bond\n",
      "b_rating\n",
      "a_rating\n",
      "aaa_rating\n",
      "aa_rating\n",
      "bbb_rating\n",
      "duration_bond\n",
      "portfolio_communication_allocation\n",
      "portfolio_financial_services\n",
      "portfolio_industrials_allocation\n",
      "portfolio_tech_allocation\n",
      "portfolio_materials_basic_allocation\n",
      "portfolio_energy_allocation\n",
      "portfolio_consumer_defence_allocation\n",
      "portfolio_healthcare_allocation\n",
      "portfolio_property_allocation\n",
      "id\n",
      "portfolio_utils_allocation\n",
      "portfolio_cyclical_consumer_allocation\n",
      "fund_ratio_net_annual_expense\n",
      "pb_ratio\n",
      "total_assets\n",
      "yield\n",
      "2014_category_return\n",
      "2012_return_category\n",
      "years_up\n",
      "2018_return_category\n",
      "category_return_1year\n",
      "cash_percent_of_portfolio\n",
      "2011_return_category\n",
      "ytd_return_fund\n",
      "years_down\n",
      "2014_return_fund\n",
      "category_return_1month\n",
      "2013_return_fund\n",
      "fund_return_3months\n",
      "ytd_return_category\n",
      "2017_category_return\n",
      "1_year_return_fund\n",
      "2015_return_fund\n",
      "portfolio_convertable\n",
      "3_months_return_category\n",
      "portfolio_others\n",
      "2016_return_fund\n",
      "stock_percent_of_portfolio\n",
      "2016_return_category\n",
      "2011_return_fund\n",
      "2010_return_fund\n",
      "fund_return_3years\n",
      "2012_fund_return\n",
      "2018_return_fund\n",
      "2017_return_fund\n",
      "category_ratio_net_annual_expense\n",
      "category_return_2015\n",
      "1_month_fund_return\n",
      "bond_percentage_of_porfolio\n",
      "portfolio_preferred\n",
      "2010_return_category\n",
      "2013_category_return\n",
      "3_years_alpha_fund\n",
      "3years_category_std\n",
      "3yrs_sharpe_ratio_fund\n",
      "3yrs_treynor_ratio_category\n",
      "3_years_return_mean_annual_fund\n",
      "fund_beta_3years\n",
      "3years_fund_r_squared\n",
      "3years_fund_std\n",
      "category_beta_3years\n",
      "fund_return_3years.1\n",
      "3_years_alpha_category\n",
      "3_years_return_mean_annual_category\n",
      "3yrs_sharpe_ratio_category\n",
      "3years_category_r_squared\n",
      "3_years_return_category\n",
      "category_r_squared_5years\n",
      "5yrs_sharpe_ratio_fund\n",
      "5_years_alpha_fund\n",
      "5years_fund_r_squared\n",
      "5years_fund_std\n",
      "5yrs_sharpe_ratio_category\n",
      "5_years_beta_fund\n",
      "5_years_return_mean_annual_fund\n",
      "5_years_return_mean_annual_category\n",
      "5yrs_treynor_ratio_category\n",
      "5_years_return_fund\n",
      "5_years_alpha_category\n",
      "5_years_beta_category\n",
      "5years_category_std\n",
      "5_years_return_category\n",
      "10years_category_r_squared\n",
      "10yrs_sharpe_ratio_fund\n",
      "10_years_alpha_fund\n",
      "10years_fund_r_squared\n",
      "10years_fund_std\n",
      "10yrs_sharpe_ratio_category\n",
      "10_years_beta_fund\n",
      "10_years_return_mean_annual_category\n",
      "10yrs_treynor_ratio_category\n",
      "10_years_return_fund\n",
      "10_years_alpha_category\n",
      "10_years_beta_category\n",
      "10years_category_std\n",
      "10_years_return_mean_annual_fund\n",
      "10_years_return_category\n"
     ]
    }
   ],
   "source": [
    "def replace_outlier(FundRating_in, col_name):\n",
    "    q1 = FundRating_in[col_name].quantile(0.25)\n",
    "    q3 = FundRating_in[col_name].quantile(0.75)\n",
    "    iqr = q3-q1 #Interquartile range\n",
    "    fence_low = q1-1.5*iqr\n",
    "    fence_high = q3+1.5*iqr\n",
    "    FundRating_in[col_name]=np.where((FundRating_in[col_name] < fence_low), FundRating_in[col_name].quantile(0.25), FundRating_in[col_name])\n",
    "    FundRating_in[col_name]=np.where((FundRating_in[col_name] > fence_high), FundRating_in[col_name].quantile(0.75), FundRating_in[col_name])\n",
    "    #     FundRating_in[col_name]=np.where(FundRating_in.loc[FundRating_in[col_name] > fence_high], FundRating_in[col_name].median(),FundRating_in[col_name])\n",
    "    #median=FundRating[col_name].median()\n",
    "    #FundRating_in[col_name] = np.where((FundRating_in[col_name] > fence_high) | (FundRating_in[col_name] < fence_low), median, FundRating_in[col_name])\n",
    "    return FundRating_in\n",
    "    \n",
    "df_columns = FundRating.columns\n",
    "for i in range(df_columns.size):\n",
    "    print(df_columns[i])\n",
    "    FundRating=replace_outlier(FundRating, df_columns[i])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [],
   "source": [
    "cleanup_nums = {\n",
    "                \"greatstone_rating\": {\"zero\":0, \"one\":1 , \"two\":2, \"three\":3, \"four\":4, \"five\":5 }\n",
    "               }\n",
    "#Replacing in the data set\n",
    "FundRating1.replace(cleanup_nums, inplace=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 248,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.impute import KNNImputer\n",
    "#  #nan = np.nan\n",
    "# #>>> X = [[1, 2, nan], [3, 4, 3], [nan, 6, 5], [8, 8, 7]]\n",
    "# imputer = KNNImputer(n_neighbors=2, weights=\"uniform\")\n",
    "# imputer.fit_transform(FundRating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 249,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Split the data into train and test \n",
    "#(Suggestion: specify “random state” if you are using train_test_split from Sklearn)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 250,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(15705    4\n",
       " 6016     3\n",
       " 24764    1\n",
       " 11962    2\n",
       " 15824    2\n",
       "         ..\n",
       " 13123    4\n",
       " 19648    0\n",
       " 9845     3\n",
       " 10799    2\n",
       " 2732     3\n",
       " Name: greatstone_rating, Length: 17500, dtype: int64, 14149    5\n",
       " 8946     5\n",
       " 22378    3\n",
       " 12162    3\n",
       " 4879     4\n",
       "         ..\n",
       " 3639     3\n",
       " 3338     2\n",
       " 10927    3\n",
       " 23669    3\n",
       " 6492     2\n",
       " Name: greatstone_rating, Length: 7500, dtype: int64)"
      ]
     },
     "execution_count": 250,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Split the data into train and test \n",
    "#(Suggestion: specify “random state” if you are using train_test_split from Sklearn)\n",
    "\n",
    "#Importing required libaries\n",
    "from sklearn.model_selection import train_test_split\n",
    "import os\n",
    "\n",
    "#Drop the target 'class' from data\n",
    "#x=FundRating.drop('greatstone_rating',axis=1)\n",
    "x=FundRating\n",
    "y=FundRating1['greatstone_rating']\n",
    "# Splitting the dataset into the Training set and Test set\n",
    "x_train,x_test,y_train,y_test=train_test_split(x,y,train_size=0.70,test_size=0.30,random_state=0)\n",
    "x_train,x_test\n",
    "y_train,y_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 251,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Do data Scaling/Normalization if required\n",
    "#This is needed we have variables in different scale ex: scaled_variance.1 etc\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "sc = StandardScaler()\n",
    "x_train= sc.fit_transform(x_train)\n",
    "x_test = sc.fit_transform(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 252,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Use PCA from Scikit learn, extract Principal Components that capture about 95% of the variance in the data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 253,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # generating the covariance matrix and the eigen values for the PCA analysis\n",
    "# cov_matrix = np.cov(x_train.T) # the relevanat covariance matrix\n",
    "# print('Covariance Matrix \\n%s', cov_matrix)\n",
    "\n",
    "# #generating the eigen values and the eigen vectors\n",
    "# e_vals, e_vecs = np.linalg.eig(cov_matrix)\n",
    "# print('Eigenvectors \\n%s' %e_vecs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 254,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print('\\nEigenvalues \\n%s' %e_vals)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 255,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # \"cumulative variance explained\" analysis \n",
    "# tot = sum(e_vals)\n",
    "# var_exp = [( i /tot ) * 100 for i in sorted(e_vals, reverse=True)]\n",
    "# cum_var_exp = np.cumsum(var_exp)\n",
    "# print(\"Cumulative Variance Explained\", cum_var_exp)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 256,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # Plotting the variance expalained by the principal components and the cumulative variance explained.\n",
    "# plt.figure(figsize=(10 , 5))\n",
    "# plt.bar(range(1, e_vals.size + 1), var_exp, alpha = 0.5, align = 'center', label = 'Individual explained variance')\n",
    "# plt.step(range(1, e_vals.size + 1), cum_var_exp, where='mid', label = 'Cumulative explained variance')\n",
    "# plt.ylabel('Explained Variance Ratio')\n",
    "# plt.xlabel('Principal Components')\n",
    "# plt.legend(loc = 'best')\n",
    "# plt.tight_layout()\n",
    "# plt.show()\n",
    "\n",
    "# #Inference: From below graph we can learn that ~95% data variance can be captured \n",
    "# #when we take only ~ 60 variables from the original data set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 257,
   "metadata": {},
   "outputs": [],
   "source": [
    "# #Proceeding with 6 components\n",
    "# pca_new = PCA(n_components=95)\n",
    "# #Fitting PCA into train data\n",
    "# pca_new.fit(x_train)\n",
    "# #Fitting PCA into test data\n",
    "# pca_new.fit(x_test)\n",
    "# print(pca_new.components_)\n",
    "# print(pca_new.explained_variance_ratio_)\n",
    "# Xpca_train_new = pca_new.transform(x_train)\n",
    "# Xpca_test_new = pca_new.transform(x_test)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 258,
   "metadata": {},
   "outputs": [],
   "source": [
    "# print(\"original shape:   \", x_test.shape)\n",
    "# print(\"transformed shape:\", Xpca_test_new.shape)\n",
    "# #Inference: From below data it is evident that dimension of the data set is reduced from 18 to 6 \n",
    "# #(excluding target variable-class)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 259,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Repeat steps 3,4 and 5 but this time, use Principal Components instead of the original data. \n",
    "#And the accuracy score should be on the same rows of test data that were used earlier."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 260,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9552571428571428, 0.6476)"
      ]
     },
     "execution_count": 260,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#Before PCA-RFC\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "# param_dist = {\"n_estimators\":[55],\n",
    "#               \"max_depth\":[None],\n",
    "#               \"max_features\": sp_randint(40,45),\n",
    "#               \"min_samples_split\": [16],\n",
    "#               \"min_samples_leaf\": [5],\n",
    "#               \"bootstrap\": [True, False],\n",
    "#               \"criterion\": [ \"entropy\"]}\n",
    "#rfcl = RandomForestClassifier(n_estimators=55)\n",
    "rfcl = RandomForestClassifier(n_estimators=55,max_depth=None,max_features=44,min_samples_split=16,min_samples_leaf=5,bootstrap= False,criterion='entropy')\n",
    "#rfcl = RandomForestClassifier('bootstrap': False,'criterion':'entropy','max_depth': None,'max_features': 44,'min_samples_leaf': 5,'min_samples_split':16,'n_estimators':55)\n",
    "rfcl = rfcl.fit(x_train, y_train)\n",
    "rfcl.score(x_train, y_train), rfcl.score(x_test, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 228,
   "metadata": {},
   "outputs": [
    {
     "ename": "ValueError",
     "evalue": "Input contains NaN, infinity or a value too large for dtype('float32').",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mValueError\u001b[0m                                Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-228-204809e5b403>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mgreatstone_rating\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mrfcl\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py\u001b[0m in \u001b[0;36mpredict\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    543\u001b[0m             \u001b[0mThe\u001b[0m \u001b[0mpredicted\u001b[0m \u001b[0mclasses\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    544\u001b[0m         \"\"\"\n\u001b[0;32m--> 545\u001b[0;31m         \u001b[0mproba\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict_proba\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    546\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    547\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mn_outputs_\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py\u001b[0m in \u001b[0;36mpredict_proba\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    586\u001b[0m         \u001b[0mcheck_is_fitted\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m'estimators_'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    587\u001b[0m         \u001b[0;31m# Check data\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 588\u001b[0;31m         \u001b[0mX\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_validate_X_predict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    589\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    590\u001b[0m         \u001b[0;31m# Assign chunk of trees to jobs\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py\u001b[0m in \u001b[0;36m_validate_X_predict\u001b[0;34m(self, X)\u001b[0m\n\u001b[1;32m    357\u001b[0m                                  \"call `fit` before exploiting the model.\")\n\u001b[1;32m    358\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 359\u001b[0;31m         \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mestimators_\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_validate_X_predict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcheck_input\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mTrue\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    360\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    361\u001b[0m     \u001b[0;34m@\u001b[0m\u001b[0mproperty\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/tree/tree.py\u001b[0m in \u001b[0;36m_validate_X_predict\u001b[0;34m(self, X, check_input)\u001b[0m\n\u001b[1;32m    389\u001b[0m         \u001b[0;34m\"\"\"Validate X whenever one tries to predict, apply, predict_proba\"\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    390\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mcheck_input\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 391\u001b[0;31m             \u001b[0mX\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mcheck_array\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mX\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mDTYPE\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0maccept_sparse\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m\"csr\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    392\u001b[0m             if issparse(X) and (X.indices.dtype != np.intc or\n\u001b[1;32m    393\u001b[0m                                 X.indptr.dtype != np.intc):\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py\u001b[0m in \u001b[0;36mcheck_array\u001b[0;34m(array, accept_sparse, accept_large_sparse, dtype, order, copy, force_all_finite, ensure_2d, allow_nd, ensure_min_samples, ensure_min_features, warn_on_dtype, estimator)\u001b[0m\n\u001b[1;32m    540\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mforce_all_finite\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    541\u001b[0m             _assert_all_finite(array,\n\u001b[0;32m--> 542\u001b[0;31m                                allow_nan=force_all_finite == 'allow-nan')\n\u001b[0m\u001b[1;32m    543\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    544\u001b[0m     \u001b[0;32mif\u001b[0m \u001b[0mensure_min_samples\u001b[0m \u001b[0;34m>\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/opt/anaconda3/lib/python3.7/site-packages/sklearn/utils/validation.py\u001b[0m in \u001b[0;36m_assert_all_finite\u001b[0;34m(X, allow_nan)\u001b[0m\n\u001b[1;32m     54\u001b[0m                 not allow_nan and not np.isfinite(X).all()):\n\u001b[1;32m     55\u001b[0m             \u001b[0mtype_err\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m'infinity'\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mallow_nan\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;34m'NaN, infinity'\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 56\u001b[0;31m             \u001b[0;32mraise\u001b[0m \u001b[0mValueError\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmsg_err\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mformat\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtype_err\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     57\u001b[0m     \u001b[0;31m# for object dtype data, we only check for NaNs (GH-13254)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     58\u001b[0m     \u001b[0;32melif\u001b[0m \u001b[0mX\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdtype\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'object'\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mand\u001b[0m \u001b[0;32mnot\u001b[0m \u001b[0mallow_nan\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mValueError\u001b[0m: Input contains NaN, infinity or a value too large for dtype('float32')."
     ]
    }
   ],
   "source": [
    "\n",
    "greatstone_rating=rfcl.predict(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 225,
   "metadata": {},
   "outputs": [],
   "source": [
    "greatstone_rating"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 217,
   "metadata": {},
   "outputs": [],
   "source": [
    "greatstone_rating = pd.DataFrame(greatstone_rating, columns=['greatstone_rating']).to_csv('sample_submissionRFCL3.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 162,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0.9845714285714285, 0.6222666666666666)"
      ]
     },
     "execution_count": 162,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#After PCA-RFC\n",
    "rfcl.fit(Xpca_train_new, y_train)\n",
    "rfcl.score(Xpca_train_new, y_train), rfcl.score(Xpca_test_new, y_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'bootstrap': True,\n",
       " 'class_weight': None,\n",
       " 'criterion': 'gini',\n",
       " 'max_depth': None,\n",
       " 'max_features': 'auto',\n",
       " 'max_leaf_nodes': None,\n",
       " 'min_impurity_decrease': 0.0,\n",
       " 'min_impurity_split': None,\n",
       " 'min_samples_leaf': 1,\n",
       " 'min_samples_split': 2,\n",
       " 'min_weight_fraction_leaf': 0.0,\n",
       " 'n_estimators': 48,\n",
       " 'n_jobs': None,\n",
       " 'oob_score': False,\n",
       " 'random_state': None,\n",
       " 'verbose': 0,\n",
       " 'warm_start': False}"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfcl.get_params()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.preprocessing import StandardScaler \n",
    "from sklearn.decomposition import PCA \n",
    "from sklearn.pipeline import Pipeline \n",
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "from scipy.stats import randint as sp_randint\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "param_dist = {\"n_estimators\":[55],\n",
    "              \"max_depth\":[None],\n",
    "              \"max_features\": sp_randint(40,45),\n",
    "              \"min_samples_split\": [16],\n",
    "              \"min_samples_leaf\": [5],\n",
    "              \"bootstrap\": [True, False],\n",
    "              \"criterion\": [ \"entropy\"]}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "# run randomized search\n",
    "samples = 10  # number of random samples \n",
    "randomCV = RandomizedSearchCV(rfcl, param_distributions=param_dist, n_iter=samples,verbose=10) #default cv = 3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/chai3/opt/anaconda3/lib/python3.7/site-packages/sklearn/model_selection/_split.py:1978: FutureWarning: The default value of cv will change from 3 to 5 in version 0.22. Specify it explicitly to silence this warning.\n",
      "  warnings.warn(CV_WARNING, FutureWarning)\n",
      "[Parallel(n_jobs=1)]: Using backend SequentialBackend with 1 concurrent workers.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Fitting 3 folds for each of 10 candidates, totalling 30 fits\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.666, total=  18.9s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   1 out of   1 | elapsed:   18.9s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.674, total=  19.0s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   2 out of   2 | elapsed:   38.0s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.664, total=  19.1s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   3 out of   3 | elapsed:   57.1s remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.662, total=  12.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   4 out of   4 | elapsed:  1.2min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.656, total=  11.9s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   5 out of   5 | elapsed:  1.3min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.652, total=  12.1s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   6 out of   6 | elapsed:  1.6min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.663, total=  12.5s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   7 out of   7 | elapsed:  1.8min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.659, total=  12.6s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   8 out of   8 | elapsed:  2.0min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=42, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.650, total=  12.6s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done   9 out of   9 | elapsed:  2.2min remaining:    0.0s\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.668, total=  20.3s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.674, total=  20.5s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.665, total=  20.6s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.659, total=  13.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.664, total=  12.9s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=43, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.651, total=  12.9s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.657, total=  12.8s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.656, total=  12.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.655, total=  12.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.659, total=  13.2s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.661, total=  13.2s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=44, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.649, total=  14.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.654, total=  13.3s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.661, total=  15.0s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=40, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.653, total=  13.6s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.661, total=  14.5s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.659, total=  14.2s\n",
      "[CV] bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=True, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.655, total=  14.9s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.665, total=  19.8s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.674, total=  19.9s\n",
      "[CV] bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55 \n",
      "[CV]  bootstrap=False, criterion=entropy, max_depth=None, max_features=41, min_samples_leaf=5, min_samples_split=16, n_estimators=55, score=0.664, total=  20.2s\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[Parallel(n_jobs=1)]: Done  30 out of  30 | elapsed:  7.6min finished\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'bootstrap': False, 'criterion': 'entropy', 'max_depth': None, 'max_features': 44, 'min_samples_leaf': 5, 'min_samples_split': 16, 'n_estimators': 55}\n"
     ]
    }
   ],
   "source": [
    "randomCV.fit(x_train, y_train)\n",
    " \n",
    "print(randomCV.best_params_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.67\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'entropy', 'max_depth': None, 'max_features': 44, 'min_samples_leaf': 5, 'min_samples_split': 16, 'n_estimators': 55}\n",
      " Test set accuracy: 0.64\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 158,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.52\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'entropy', 'max_depth': 5, 'max_features': 41, 'min_samples_leaf': 5, 'min_samples_split': 16, 'n_estimators': 68}\n",
      " Test set accuracy: 0.47\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 145,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.67\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'entropy', 'max_depth': None, 'max_features': 43, 'min_samples_leaf': 5, 'min_samples_split': 16, 'n_estimators': 55}\n",
      " Test set accuracy: 0.64\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 141,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.66\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'gini', 'max_depth': None, 'max_features': 46, 'min_samples_leaf': 3, 'min_samples_split': 17, 'n_estimators': 70}\n",
      " Test set accuracy: 0.63\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 137,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.66\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'entropy', 'max_depth': None, 'max_features': 43, 'min_samples_leaf': 9, 'min_samples_split': 27, 'n_estimators': 40}\n",
      " Test set accuracy: 0.63\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 128,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      " Best cross-validation accuracy: 0.67\n",
      " Best parameters:  {'bootstrap': False, 'criterion': 'entropy', 'max_depth': None, 'max_features': 37, 'min_samples_leaf': 5, 'min_samples_split': 16, 'n_estimators': 55}\n",
      " Test set accuracy: 0.64\n"
     ]
    }
   ],
   "source": [
    "print(\" Best cross-validation accuracy: {:.2f}\". format( randomCV.best_score_)) \n",
    "print(\" Best parameters: \", randomCV.best_params_) \n",
    "print(\" Test set accuracy: {:.2f}\". format( randomCV.score( x_test, y_test)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 117,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from sklearn.model_selection import GridSearchCV\n",
    "# #pipe_svc = Pipeline([('scl', StandardScaler()), ('pca', PCA()), ('svc', SVC())]) \n",
    "# pipe_svc = Pipeline([('scl', StandardScaler()),('rfcl', RandomForestClassifier())]) \n",
    "# #param_grid = {'pca__n_components':[65],'svc__C': [0.001, 0.01, 0.1, 1,5,6,7,8,9,10], 'svc__gamma': [0.001,0.01,0.06,0.5,0.1,1,2,3,4,10], 'svc__kernel':['rbf','poly']} \n",
    "\n",
    "# param_grid = {'svc__C': [5,10,14], 'svc__gamma': [0.05,0.06,0.07,], 'svc__kernel':['rbf']} \n",
    "\n",
    "# grid = GridSearchCV( pipe_svc , param_grid = param_grid, cv = 5,verbose=10) \n",
    "\n",
    "# grid.fit( x_train, y_train) \n",
    "\n",
    "# print(\" Best cross-validation accuracy: {:.2f}\". format( grid.best_score_)) \n",
    "# print(\" Best parameters: \", grid.best_params_) \n",
    "# print(\" Test set accuracy: {:.2f}\". format( grid.score( x_cv, y_cv)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 129,
   "metadata": {},
   "outputs": [],
   "source": [
    "greatstone_rating = rfcl.predict(x_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 130,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[5 5 4 ... 3 3 2]\n"
     ]
    }
   ],
   "source": [
    "print (greatstone_rating)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 131,
   "metadata": {},
   "outputs": [],
   "source": [
    "greatstone_rating = pd.DataFrame(greatstone_rating, columns=['greatstone_rating']).to_csv('sample_submissionRFCL.csv')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.7"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}