Abalone (1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Predicting the Age of Abalone From Physical Measurements"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Frame the problem and look at the big picture:\n",
    "\n",
    "#### In this dataset, we must create a model that streamlines the process of determining the age of abalone. A more invasive and tedious approach to determining the age of an abalone is to disect it from the shell and count the rings within it. The number of rings directly correlates to the age. However, there are other physical features that are indicative of the abalone's age, and those features are to be used for training our model to predict the age of in the abalone. Essentially, this model will reduce the need to harm the abalone in order to collect data on it and monitor the growth of marine life.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Get the data:\n",
    "\n",
    "#### In this dataset, we will read in the file as a csv in pandas and correctly label the columns as the features that are defined here: http://archive.ics.uci.edu/ml/datasets/Abalone. \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv(\"/Users/ryanmartin/Desktop/abalone (3).data\", header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>0</th>\n",
       "      <th>1</th>\n",
       "      <th>2</th>\n",
       "      <th>3</th>\n",
       "      <th>4</th>\n",
       "      <th>5</th>\n",
       "      <th>6</th>\n",
       "      <th>7</th>\n",
       "      <th>8</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>M</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>F</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   0      1      2      3       4       5       6      7   8\n",
       "0  M  0.455  0.365  0.095  0.5140  0.2245  0.1010  0.150  15\n",
       "1  M  0.350  0.265  0.090  0.2255  0.0995  0.0485  0.070   7\n",
       "2  F  0.530  0.420  0.135  0.6770  0.2565  0.1415  0.210   9\n",
       "3  M  0.440  0.365  0.125  0.5160  0.2155  0.1140  0.155  10\n",
       "4  I  0.330  0.255  0.080  0.2050  0.0895  0.0395  0.055   7"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.rename(columns={0:\"Sex\", 1:\"Length\", 2: \"Diameter\", 3: \"Height\", 4: \"Whole Weight\", 5: \"Shucked Weight\",\n",
    "                  6: \"Viscera Weight\", 7: \"Shell Weight\", 8: \"Rings\"})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Viscera Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>Rings</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>M</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>F</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>7</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Sex  Length  Diameter  Height  Whole Weight  Shucked Weight  Viscera Weight  \\\n",
       "0   M   0.455     0.365   0.095        0.5140          0.2245          0.1010   \n",
       "1   M   0.350     0.265   0.090        0.2255          0.0995          0.0485   \n",
       "2   F   0.530     0.420   0.135        0.6770          0.2565          0.1415   \n",
       "3   M   0.440     0.365   0.125        0.5160          0.2155          0.1140   \n",
       "4   I   0.330     0.255   0.080        0.2050          0.0895          0.0395   \n",
       "\n",
       "   Shell Weight  Rings  \n",
       "0         0.150     15  \n",
       "1         0.070      7  \n",
       "2         0.210      9  \n",
       "3         0.155     10  \n",
       "4         0.055      7  "
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### From the problem statement and feature discription,  we must compute the target variable of the problem ' Age' and assign it to the dataset. Age = 1.5 + Rings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    " df['Age'] = df['Rings']+1.5\n",
    " df.drop('Rings', axis = 1, inplace = True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Viscera Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>M</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>16.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>M</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>F</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>10.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>M</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>11.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>I</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "  Sex  Length  Diameter  Height  Whole Weight  Shucked Weight  Viscera Weight  \\\n",
       "0   M   0.455     0.365   0.095        0.5140          0.2245          0.1010   \n",
       "1   M   0.350     0.265   0.090        0.2255          0.0995          0.0485   \n",
       "2   F   0.530     0.420   0.135        0.6770          0.2565          0.1415   \n",
       "3   M   0.440     0.365   0.125        0.5160          0.2155          0.1140   \n",
       "4   I   0.330     0.255   0.080        0.2050          0.0895          0.0395   \n",
       "\n",
       "   Shell Weight   Age  \n",
       "0         0.150  16.5  \n",
       "1         0.070   8.5  \n",
       "2         0.210  10.5  \n",
       "3         0.155  11.5  \n",
       "4         0.055   8.5  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Explore the data:\n",
    "\n",
    "#### We will then explore the data by studying its attributes and characteristics. By calling the pandas info function we can see all the datatypes present in each feature in the data. This information will be useful during the feature engineering stage, where we must prepare the data for the machine learning models. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 4177 entries, 0 to 4176\n",
      "Data columns (total 9 columns):\n",
      "Sex               4177 non-null object\n",
      "Length            4177 non-null float64\n",
      "Diameter          4177 non-null float64\n",
      "Height            4177 non-null float64\n",
      "Whole Weight      4177 non-null float64\n",
      "Shucked Weight    4177 non-null float64\n",
      "Viscera Weight    4177 non-null float64\n",
      "Shell Weight      4177 non-null float64\n",
      "Age               4177 non-null float64\n",
      "dtypes: float64(8), object(1)\n",
      "memory usage: 293.8+ KB\n"
     ]
    }
   ],
   "source": [
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Prepare the data:\n",
    "\n",
    "#### Data preparation for this dataset will not be too intensive because we do not have any missing values.  However, we must convert the Sex feature from categorical to numeric data in order for the machine learning algorithm to use this information as indicator variables. This will add n columns to the dataframe where n is the number of classes within that feature. In our case, since we have 3 different classes in Sex (Male, Female, and Intersex), we will be adding three new columns because the function labels the instance as a 1 if it belongs to the class and 0 if it does not. The next 4 cells displays how this process occured, and In[8] shows how the columns were reorganized to keep the label we are predicting on the end of the dataframe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Viscera Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>Age</th>\n",
       "      <th>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.1010</td>\n",
       "      <td>0.150</td>\n",
       "      <td>16.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.0485</td>\n",
       "      <td>0.070</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.1415</td>\n",
       "      <td>0.210</td>\n",
       "      <td>10.5</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.1140</td>\n",
       "      <td>0.155</td>\n",
       "      <td>11.5</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.0395</td>\n",
       "      <td>0.055</td>\n",
       "      <td>8.5</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Length  Diameter  Height  Whole Weight  Shucked Weight  Viscera Weight  \\\n",
       "0   0.455     0.365   0.095        0.5140          0.2245          0.1010   \n",
       "1   0.350     0.265   0.090        0.2255          0.0995          0.0485   \n",
       "2   0.530     0.420   0.135        0.6770          0.2565          0.1415   \n",
       "3   0.440     0.365   0.125        0.5160          0.2155          0.1140   \n",
       "4   0.330     0.255   0.080        0.2050          0.0895          0.0395   \n",
       "\n",
       "   Shell Weight   Age  Sex_F  Sex_I  Sex_M  \n",
       "0         0.150  16.5      0      0      1  \n",
       "1         0.070   8.5      0      0      1  \n",
       "2         0.210  10.5      1      0      0  \n",
       "3         0.155  11.5      0      0      1  \n",
       "4         0.055   8.5      0      1      0  "
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = pd.get_dummies(df)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "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>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.150</td>\n",
       "      <td>16.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.070</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.210</td>\n",
       "      <td>10.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.155</td>\n",
       "      <td>11.5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.055</td>\n",
       "      <td>8.5</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Sex_F  Sex_I  Sex_M  Length  Diameter  Height  Whole Weight  \\\n",
       "0      0      0      1   0.455     0.365   0.095        0.5140   \n",
       "1      0      0      1   0.350     0.265   0.090        0.2255   \n",
       "2      1      0      0   0.530     0.420   0.135        0.6770   \n",
       "3      0      0      1   0.440     0.365   0.125        0.5160   \n",
       "4      0      1      0   0.330     0.255   0.080        0.2050   \n",
       "\n",
       "   Shucked Weight  Shell Weight   Age  \n",
       "0          0.2245         0.150  16.5  \n",
       "1          0.0995         0.070   8.5  \n",
       "2          0.2565         0.210  10.5  \n",
       "3          0.2155         0.155  11.5  \n",
       "4          0.0895         0.055   8.5  "
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df = df[['Sex_F', 'Sex_I', 'Sex_M','Length', 'Diameter', 'Height', 'Whole Weight',\n",
    "        'Shucked Weight', 'Shell Weight', 'Age']]\n",
    "\n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### More data exploration:\n",
    "\n",
    "#### The describe function in pandas gives us a lot of useful statistical information on the dataset in regards to the distribution, the means, and standard deviation. And the value counts function below shows how many instances belong to that certain class. In our data, most abalone are labelled as having an age of 10.5 years. It seems that very few live past the age of 25. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>Age</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "      <td>4177.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>0.312904</td>\n",
       "      <td>0.321283</td>\n",
       "      <td>0.365813</td>\n",
       "      <td>0.523992</td>\n",
       "      <td>0.407881</td>\n",
       "      <td>0.139516</td>\n",
       "      <td>0.828742</td>\n",
       "      <td>0.359367</td>\n",
       "      <td>0.238831</td>\n",
       "      <td>11.433684</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>0.463731</td>\n",
       "      <td>0.467025</td>\n",
       "      <td>0.481715</td>\n",
       "      <td>0.120093</td>\n",
       "      <td>0.099240</td>\n",
       "      <td>0.041827</td>\n",
       "      <td>0.490389</td>\n",
       "      <td>0.221963</td>\n",
       "      <td>0.139203</td>\n",
       "      <td>3.224169</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.075000</td>\n",
       "      <td>0.055000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.002000</td>\n",
       "      <td>0.001000</td>\n",
       "      <td>0.001500</td>\n",
       "      <td>2.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.450000</td>\n",
       "      <td>0.350000</td>\n",
       "      <td>0.115000</td>\n",
       "      <td>0.441500</td>\n",
       "      <td>0.186000</td>\n",
       "      <td>0.130000</td>\n",
       "      <td>9.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.545000</td>\n",
       "      <td>0.425000</td>\n",
       "      <td>0.140000</td>\n",
       "      <td>0.799500</td>\n",
       "      <td>0.336000</td>\n",
       "      <td>0.234000</td>\n",
       "      <td>10.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.615000</td>\n",
       "      <td>0.480000</td>\n",
       "      <td>0.165000</td>\n",
       "      <td>1.153000</td>\n",
       "      <td>0.502000</td>\n",
       "      <td>0.329000</td>\n",
       "      <td>12.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.815000</td>\n",
       "      <td>0.650000</td>\n",
       "      <td>1.130000</td>\n",
       "      <td>2.825500</td>\n",
       "      <td>1.488000</td>\n",
       "      <td>1.005000</td>\n",
       "      <td>30.500000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             Sex_F        Sex_I        Sex_M       Length     Diameter  \\\n",
       "count  4177.000000  4177.000000  4177.000000  4177.000000  4177.000000   \n",
       "mean      0.312904     0.321283     0.365813     0.523992     0.407881   \n",
       "std       0.463731     0.467025     0.481715     0.120093     0.099240   \n",
       "min       0.000000     0.000000     0.000000     0.075000     0.055000   \n",
       "25%       0.000000     0.000000     0.000000     0.450000     0.350000   \n",
       "50%       0.000000     0.000000     0.000000     0.545000     0.425000   \n",
       "75%       1.000000     1.000000     1.000000     0.615000     0.480000   \n",
       "max       1.000000     1.000000     1.000000     0.815000     0.650000   \n",
       "\n",
       "            Height  Whole Weight  Shucked Weight  Shell Weight          Age  \n",
       "count  4177.000000   4177.000000     4177.000000   4177.000000  4177.000000  \n",
       "mean      0.139516      0.828742        0.359367      0.238831    11.433684  \n",
       "std       0.041827      0.490389        0.221963      0.139203     3.224169  \n",
       "min       0.000000      0.002000        0.001000      0.001500     2.500000  \n",
       "25%       0.115000      0.441500        0.186000      0.130000     9.500000  \n",
       "50%       0.140000      0.799500        0.336000      0.234000    10.500000  \n",
       "75%       0.165000      1.153000        0.502000      0.329000    12.500000  \n",
       "max       1.130000      2.825500        1.488000      1.005000    30.500000  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "10.5    689\n",
       "11.5    634\n",
       "9.5     568\n",
       "12.5    487\n",
       "8.5     391\n",
       "13.5    267\n",
       "7.5     259\n",
       "14.5    203\n",
       "15.5    126\n",
       "6.5     115\n",
       "16.5    103\n",
       "17.5     67\n",
       "18.5     58\n",
       "5.5      57\n",
       "19.5     42\n",
       "20.5     32\n",
       "21.5     26\n",
       "4.5      15\n",
       "22.5     14\n",
       "24.5      9\n",
       "23.5      6\n",
       "25.5      2\n",
       "28.5      2\n",
       "3.5       1\n",
       "30.5      1\n",
       "2.5       1\n",
       "27.5      1\n",
       "26.5      1\n",
       "Name: Age, dtype: int64"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df['Age'].value_counts()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Just checking to make sure we are indeed not missing any data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Sex_F             0\n",
       "Sex_I             0\n",
       "Sex_M             0\n",
       "Length            0\n",
       "Diameter          0\n",
       "Height            0\n",
       "Whole Weight      0\n",
       "Shucked Weight    0\n",
       "Shell Weight      0\n",
       "Age               0\n",
       "dtype: int64"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.isnull().sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Create a Test Set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Since we are predicting the age of the abalone, and the age is directly correlated to the number of rings, we will create our test set by setting the number of rings equal to y and the rest of the features as X. However, this process will officially be done later in the notebook with train/test split. This cell just demonstrates how the data will be split. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "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>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.150</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.070</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.210</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.155</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.055</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Sex_F  Sex_I  Sex_M  Length  Diameter  Height  Whole Weight  \\\n",
       "0      0      0      1   0.455     0.365   0.095        0.5140   \n",
       "1      0      0      1   0.350     0.265   0.090        0.2255   \n",
       "2      1      0      0   0.530     0.420   0.135        0.6770   \n",
       "3      0      0      1   0.440     0.365   0.125        0.5160   \n",
       "4      0      1      0   0.330     0.255   0.080        0.2050   \n",
       "\n",
       "   Shucked Weight  Shell Weight  \n",
       "0          0.2245         0.150  \n",
       "1          0.0995         0.070  \n",
       "2          0.2565         0.210  \n",
       "3          0.2155         0.155  \n",
       "4          0.0895         0.055  "
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y = df[\"Age\"]\n",
    "X = df.drop([\"Age\"], axis = 1)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Discover and Visualize Data to Gain Insights:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 9 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "import matplotlib.pyplot as plt\n",
    "df.drop([\"Sex_F\",\"Sex_I\",\"Sex_M\"], axis=1).hist(bins=50, figsize=(20,15))\n",
    "plt.show();"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### This graph shows that as the abalone grows in diameter, the whole weight grows exponentially. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot(kind=\"scatter\", x =\"Diameter\", y= \"Whole Weight\");"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### By calculating the correlation between the features and age, we can see that Shell Weight, Diameter, Height, and Length are the most positively correlated features. The sex of the abalone is less correlated, but it is interesting to see that the female is most positively correlated, probably because they live longer than the males. The intersexed abalone is the only feature with a negative correlation, which means that they do not survive as long as the males and females.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "corr_matrix = df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age               1.000000\n",
       "Shell Weight      0.627574\n",
       "Diameter          0.574660\n",
       "Height            0.557467\n",
       "Length            0.556720\n",
       "Whole Weight      0.540390\n",
       "Shucked Weight    0.420884\n",
       "Sex_F             0.250279\n",
       "Sex_M             0.181831\n",
       "Sex_I            -0.436063\n",
       "Name: Age, dtype: float64"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix['Age'].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x576 with 16 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "from pandas.plotting import scatter_matrix\n",
    "\n",
    "attributes = [\"Age\", \"Shell Weight\", \"Diameter\",\n",
    "              \"Height\"]\n",
    "scatter_matrix(df[attributes], figsize=(12, 8));"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The graph below reveals that there is a strong correlation between the diameter of abalone and its age. There is an upward trend in the graph and the points are not too dispersed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df.plot(kind=\"scatter\", x=\"Diameter\", y=\"Age\",\n",
    "             alpha=0.1)\n",
    "plt.axis([0, .7, 0, 30]);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Attribute Combinations can increase the interestingess of our dataset."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"Abalone_Mass_Index\"] = df[\"Shell Weight\"]/ df[\"Diameter\"]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### According to the correlation table, the \"Abalone Mass Index\", which is the the shell weight/ diameter, is highly correlated to the number of rings. In fact, it is more correlated than any feature in the dataset. Therefore, we will add this attribute to our dataframe to use for training our model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "corr_matrix = df.corr()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Age                   1.000000\n",
       "Abalone_Mass_Index    0.663056\n",
       "Shell Weight          0.627574\n",
       "Diameter              0.574660\n",
       "Height                0.557467\n",
       "Length                0.556720\n",
       "Whole Weight          0.540390\n",
       "Shucked Weight        0.420884\n",
       "Sex_F                 0.250279\n",
       "Sex_M                 0.181831\n",
       "Sex_I                -0.436063\n",
       "Name: Age, dtype: float64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "corr_matrix[\"Age\"].sort_values(ascending=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df.drop([\"Abalone_Mass_Index\"], axis=1)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Prepare the Data for Machine Learning Algorithms"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Because we already changed our string values to numeric values with get_dummies, and there are no missing values, we do not have too much work to do in preparing our data for machine learning models. We will add a feature to our dataset using a function transformer that will take our training data and allow us to create a function that will use the information we have to create new data. In our case, we will use the shell weight and diameter features to create the abalone mass index feature we explored in attribute combinations. This feature will then be added to our dataframe. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.base import BaseEstimator, TransformerMixin\n",
    "\n",
    "# get the right column indices: safer than hard-coding indices 3, 4, 5, 6\n",
    "ShellWeight_ix, Diameter_ix = [\n",
    "    list(df.columns).index(col)\n",
    "    for col in (\"Shell Weight\", \"Diameter\")]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from sklearn.preprocessing import FunctionTransformer\n",
    "\n",
    "def add_extra_features(X, add_abalone_mass_index=True):\n",
    "    abalone_mass_index = X[:, ShellWeight_ix]/ X[:, Diameter_ix]\n",
    "    return np.c_[X, abalone_mass_index]\n",
    "\n",
    "attr_adder = FunctionTransformer(add_extra_features, validate=False,\n",
    "                                 kw_args={\"add_abalone_mass_index\": False})\n",
    "abalone_extra_attribs = attr_adder.fit_transform(df.values)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "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>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>abalone_mass_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.150</td>\n",
       "      <td>0.410959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.264151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.210</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.155</td>\n",
       "      <td>0.424658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.215686</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Sex_F  Sex_I  Sex_M  Length  Diameter  Height  Whole Weight  \\\n",
       "0    0.0    0.0    1.0   0.455     0.365   0.095        0.5140   \n",
       "1    0.0    0.0    1.0   0.350     0.265   0.090        0.2255   \n",
       "2    1.0    0.0    0.0   0.530     0.420   0.135        0.6770   \n",
       "3    0.0    0.0    1.0   0.440     0.365   0.125        0.5160   \n",
       "4    0.0    1.0    0.0   0.330     0.255   0.080        0.2050   \n",
       "\n",
       "   Shucked Weight  Shell Weight  abalone_mass_index  \n",
       "0          0.2245         0.150            0.410959  \n",
       "1          0.0995         0.070            0.264151  \n",
       "2          0.2565         0.210            0.500000  \n",
       "3          0.2155         0.155            0.424658  \n",
       "4          0.0895         0.055            0.215686  "
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abalone_extra_attribs = pd.DataFrame(\n",
    "    abalone_extra_attribs,\n",
    "    columns=list(df.columns)+[\"abalone_mass_index\"],\n",
    "    index=df.index)\n",
    "\n",
    "X = abalone_extra_attribs.drop([\"Age\"], axis = 1)\n",
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will  then use a standard scaler for Standarization, which will subtract the mean value and then divide it by the standard deviation so that the resulting distribution has unit variance. This will lower the effectiveness of outliers. We will print out the example transformed data just for demonstration, but this pipeline will offically fit our model in a later stage. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.feature_selection import SelectFromModel\n",
    "from sklearn.svm import LinearSVC\n",
    "\n",
    "full_pipeline = pipeline = Pipeline([\n",
    "    ('std_scaler', StandardScaler())\n",
    "])\n",
    "\n",
    "abalone_prepared = full_pipeline.fit_transform(df)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.67483383, -0.68801788,  1.31667716, ..., -0.60768536,\n",
       "        -0.63821689,  1.57154357],\n",
       "       [-0.67483383, -0.68801788,  1.31667716, ..., -1.17090984,\n",
       "        -1.21298732, -0.91001299],\n",
       "       [ 1.48184628, -0.68801788, -0.75948762, ..., -0.4634999 ,\n",
       "        -0.20713907, -0.28962385],\n",
       "       ...,\n",
       "       [-0.67483383, -0.68801788,  1.31667716, ...,  0.74855917,\n",
       "         0.49695471, -0.28962385],\n",
       "       [ 1.48184628, -0.68801788, -0.75948762, ...,  0.77334105,\n",
       "         0.41073914,  0.02057072],\n",
       "       [-0.67483383, -0.68801788,  1.31667716, ...,  2.64099341,\n",
       "         1.84048058,  0.64095986]])"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abalone_prepared"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4177, 10)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "abalone_prepared.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "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>Sex_F</th>\n",
       "      <th>Sex_I</th>\n",
       "      <th>Sex_M</th>\n",
       "      <th>Length</th>\n",
       "      <th>Diameter</th>\n",
       "      <th>Height</th>\n",
       "      <th>Whole Weight</th>\n",
       "      <th>Shucked Weight</th>\n",
       "      <th>Shell Weight</th>\n",
       "      <th>abalone_mass_index</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.455</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.095</td>\n",
       "      <td>0.5140</td>\n",
       "      <td>0.2245</td>\n",
       "      <td>0.150</td>\n",
       "      <td>0.410959</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.350</td>\n",
       "      <td>0.265</td>\n",
       "      <td>0.090</td>\n",
       "      <td>0.2255</td>\n",
       "      <td>0.0995</td>\n",
       "      <td>0.070</td>\n",
       "      <td>0.264151</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.530</td>\n",
       "      <td>0.420</td>\n",
       "      <td>0.135</td>\n",
       "      <td>0.6770</td>\n",
       "      <td>0.2565</td>\n",
       "      <td>0.210</td>\n",
       "      <td>0.500000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.440</td>\n",
       "      <td>0.365</td>\n",
       "      <td>0.125</td>\n",
       "      <td>0.5160</td>\n",
       "      <td>0.2155</td>\n",
       "      <td>0.155</td>\n",
       "      <td>0.424658</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.330</td>\n",
       "      <td>0.255</td>\n",
       "      <td>0.080</td>\n",
       "      <td>0.2050</td>\n",
       "      <td>0.0895</td>\n",
       "      <td>0.055</td>\n",
       "      <td>0.215686</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   Sex_F  Sex_I  Sex_M  Length  Diameter  Height  Whole Weight  \\\n",
       "0    0.0    0.0    1.0   0.455     0.365   0.095        0.5140   \n",
       "1    0.0    0.0    1.0   0.350     0.265   0.090        0.2255   \n",
       "2    1.0    0.0    0.0   0.530     0.420   0.135        0.6770   \n",
       "3    0.0    0.0    1.0   0.440     0.365   0.125        0.5160   \n",
       "4    0.0    1.0    0.0   0.330     0.255   0.080        0.2050   \n",
       "\n",
       "   Shucked Weight  Shell Weight  abalone_mass_index  \n",
       "0          0.2245         0.150            0.410959  \n",
       "1          0.0995         0.070            0.264151  \n",
       "2          0.2565         0.210            0.500000  \n",
       "3          0.2155         0.155            0.424658  \n",
       "4          0.0895         0.055            0.215686  "
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train Test Split"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To avoid creating a machine learning model that would just repeat the labels that it had just seen and fail to predict any new unseen data, we must divide and set aside a portion of our data as a test set (X_test, y_test). This will help avoid the situation of overfitting our model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X, y, random_state=0, test_size = 0.2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Select and Train a Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Classification and Regression"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will start our modelling experiment with Logistic Regression that supports multiclass labels (Softmax Regression). When given an instance, the Softmax Regression model will compute the score for each class, and then estimate the probablility of each class by applying the softmax function to the scores. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LogisticRegression"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "log_reg = LogisticRegression(multi_class=\"multinomial\", solver=\"lbfgs\", C=10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "lc = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', log_reg)])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will then fit our model to our training data through the pipeline we created in the cell above. The training data first is scaled with the standard scaler to reduce the influence of outliers, then it is trained with softmax regression to estimate the probability that an instance belongs to a certain class. When given an instance i, the softmax regression model first computes a score Sk(i) for each class k, and then it estimates the probability of each class by applying the softmax function."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Our classification model cannot work with continuous labels, so we will \"cheat\" and round our y values to make this model work, although it will not be very accurate. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "X_train_clf, X_test_clf, y_train_clf, y_test_clf = train_test_split(X, y.round(), random_state=0, test_size = 0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:947: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
      "  \"of iterations.\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 Pipeline(memory=None,\n",
       "                          steps=[('std_scaler',\n",
       "                                  StandardScaler(copy=True, with_mean=True,\n",
       "                                                 with_std=True))],\n",
       "                          verbose=False)),\n",
       "                ('classifier',\n",
       "                 LogisticRegression(C=10, class_weight=None, dual=False,\n",
       "                                    fit_intercept=True, intercept_scaling=1,\n",
       "                                    l1_ratio=None, max_iter=100,\n",
       "                                    multi_class='multinomial', n_jobs=None,\n",
       "                                    penalty='l2', random_state=None,\n",
       "                                    solver='lbfgs', tol=0.0001, verbose=0,\n",
       "                                    warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "lc.fit(X_train_clf, y_train_clf)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = lc.predict(X_test_clf)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We can now assess the accuracy of our model with cross validation, which will generate cross-validated estimates for each input data point. It will accomplish this task by taking the data and splitting it according to the cv parameter where each sample belongs to one test set and the prediction is computed with an estimator fitted on the training set. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/model_selection/_split.py:657: Warning: The least populated class in y has only 1 members, which is too few. The minimum number of members in any class cannot be less than n_splits=3.\n",
      "  % (min_groups, self.n_splits)), Warning)\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:947: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
      "  \"of iterations.\", ConvergenceWarning)\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:947: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
      "  \"of iterations.\", ConvergenceWarning)\n",
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/linear_model/logistic.py:947: ConvergenceWarning: lbfgs failed to converge. Increase the number of iterations.\n",
      "  \"of iterations.\", ConvergenceWarning)\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0.47942755, 0.497307  , 0.46798918])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "cross_val_score(lc, X_train_clf, y_train_clf, cv=3, scoring=\"accuracy\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Accuracy:  0.45933014354066987\n"
     ]
    }
   ],
   "source": [
    "from sklearn.metrics import accuracy_score \n",
    "accuracy = accuracy_score(y_test_clf, y_pred)\n",
    "print('Accuracy: ',accuracy)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will evaluate the weaknesses of this classification model with a confusion matrix which will show the accuracy of each classification. \n",
    "\n",
    "#### By definition, a confusion matrix shows the number of observations known to be in group but predicted to be in another group.\n",
    "\n",
    "#### Because we were attempting to predict instances belonging to many different classess, we needed to generate a multilabel confusion matrix, which provides the accuracy of each classification within each class, the abalone's age. However, because our model did poorly, we must reframe our problem and consider this as a dataset that would fit better in a regression model. We are, in fact, predicting a numeric value. Even though that numeric value represents a class, it is still a regression problem. \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import multilabel_confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[831,   0],\n",
       "        [  5,   0]],\n",
       "\n",
       "       [[777,  15],\n",
       "        [ 23,  21]],\n",
       "\n",
       "       [[630,  61],\n",
       "        [ 57,  88]],\n",
       "\n",
       "       [[430, 166],\n",
       "        [ 78, 162]],\n",
       "\n",
       "       [[467, 167],\n",
       "        [100, 102]],\n",
       "\n",
       "       [[709,  22],\n",
       "        [ 99,   6]],\n",
       "\n",
       "       [[779,  13],\n",
       "        [ 41,   3]],\n",
       "\n",
       "       [[805,   7],\n",
       "        [ 22,   2]],\n",
       "\n",
       "       [[823,   0],\n",
       "        [ 13,   0]],\n",
       "\n",
       "       [[825,   1],\n",
       "        [ 10,   0]],\n",
       "\n",
       "       [[833,   0],\n",
       "        [  3,   0]],\n",
       "\n",
       "       [[835,   0],\n",
       "        [  1,   0]]])"
      ]
     },
     "execution_count": 42,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "multilabel_confusion_matrix(y_test_clf, y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Even though we will not deploy this model for this dataset, we will compute the F1 score just to be thorough in our classification attempt. \n",
    "\n",
    "#### The F1 score takes the weighted average of the precision and recall, where an F1 score reaches its best value at 1 and worst score at 0.  \n",
    "\n",
    "#### F1 = 2 * (precision * recall) / (precision + recall)\n",
    "\n",
    "#### In our multi-class and multi-label case, we will show the average of the F1 score of each class (see classification report below)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.metrics import classification_report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "         4.0   0.000000  0.000000  0.000000         5\n",
      "         6.0   0.583333  0.477273  0.525000        44\n",
      "         8.0   0.590604  0.606897  0.598639       145\n",
      "        10.0   0.493902  0.675000  0.570423       240\n",
      "        12.0   0.379182  0.504950  0.433121       202\n",
      "        14.0   0.214286  0.057143  0.090226       105\n",
      "        16.0   0.187500  0.068182  0.100000        44\n",
      "        18.0   0.222222  0.083333  0.121212        24\n",
      "        20.0   0.000000  0.000000  0.000000        13\n",
      "        22.0   0.000000  0.000000  0.000000        10\n",
      "        24.0   0.000000  0.000000  0.000000         3\n",
      "        30.0   0.000000  0.000000  0.000000         1\n",
      "\n",
      "    accuracy                       0.459330       836\n",
      "   macro avg   0.222586  0.206065  0.203218       836\n",
      "weighted avg   0.409712  0.459330  0.419949       836\n",
      "\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/sklearn/metrics/classification.py:1437: UndefinedMetricWarning: Precision and F-score are ill-defined and being set to 0.0 in labels with no predicted samples.\n",
      "  'precision', 'predicted', average, warn_for)\n"
     ]
    }
   ],
   "source": [
    "sk_report = classification_report(\n",
    "    digits=6,\n",
    "    y_true=y_test_clf, \n",
    "    y_pred=lc.predict(X_test_clf))\n",
    "print(sk_report)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### The results of our softmax regression experiment were disappointing given that our accuracy was only .27 and the weighted average f1-score was .24. This dataset might respond better to a linear regression model that predicts a numeric output and is evaluated on a MSE. To be more precise, we will now attempt ridge regression to keep the weights as small as possible since we are dealing with many features that are highly correlated to the label. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Regression Experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.linear_model import Ridge"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "ridge_reg = Ridge(alpha=1, solver=\"cholesky\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "ridge = Pipeline(steps=[('preprocessor', full_pipeline),\n",
    "                      ('classifier', ridge_reg)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 Pipeline(memory=None,\n",
       "                          steps=[('std_scaler',\n",
       "                                  StandardScaler(copy=True, with_mean=True,\n",
       "                                                 with_std=True))],\n",
       "                          verbose=False)),\n",
       "                ('classifier',\n",
       "                 Ridge(alpha=1, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "                       normalize=False, random_state=None, solver='cholesky',\n",
       "                       tol=0.001))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ridge.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Evaluating Our Trained Model With Root Mean Square Error(RMSE) and Mean Absolute Error (MAE):\n",
    "\n",
    "The mean squared error (MSE) measures the average of the squares of the errors, which is the average squared difference between the estimated values and the actual value. The RMSE and MAE measure the distance between two vectors: the vectors of the predictions and the vector of the target values. Usually, the RMSE score will be larger in datasets that have many outlier. This is clearly happening in our case because we have a higher norm index. \n",
    "\n",
    "**Root Mean Squared Error** (RMSE) is the square root of the mean of the squared errors:\n",
    "\n",
    "$$\\sqrt{\\frac 1n\\sum_{i=1}^n(y_i-\\hat{y}_i)^2}$$\n",
    "\n",
    "**Mean Absolute Error** (MAE) is the mean of the absolute value of the errors:\n",
    "\n",
    "$$\\frac 1n\\sum_{i=1}^n|y_i-\\hat{y}_i|$$\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.1980758766383675"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_squared_error\n",
    "\n",
    "y_pred = ridge.predict(X_test)\n",
    "lin_mse = mean_squared_error(y_test, y_pred)\n",
    "lin_rmse = np.sqrt(lin_mse)\n",
    "lin_rmse"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Consequently, our MSE is a smaller value and probably more reliable because we have many outliers. (See scatterplot below. )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Predicted Y')"
      ]
     },
     "execution_count": 50,
     "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": [
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Y Test')\n",
    "plt.ylabel('Predicted Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "1.5465111654881754"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.metrics import mean_absolute_error\n",
    "\n",
    "lin_mae = mean_absolute_error(y_test, y_pred)\n",
    "lin_mae"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Our MAE score means that we can predict with our model the age of the abalone within 1.5 years of its actual age based on its physical features."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Fine Tune Our ML Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "scores = cross_val_score(ridge, X_train, y_train,\n",
    "                         scoring=\"neg_mean_squared_error\", cv=10)\n",
    "ridge_rmse_scores = np.sqrt(-scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scores: [2.25492973 1.99189974 2.07347097 2.22458661 1.9879511  2.21006217\n",
      " 2.08479332 2.43281413 2.15729603 2.24753345]\n",
      "Mean: 2.1665337244752476\n",
      "Standard deviation: 0.12975408166869484\n"
     ]
    }
   ],
   "source": [
    "def display_scores(scores):\n",
    "    print(\"Scores:\", scores)\n",
    "    print(\"Mean:\", scores.mean())\n",
    "    print(\"Standard deviation:\", scores.std())\n",
    "\n",
    "display_scores(ridge_rmse_scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### To avoid overfitting, we will perform a (supervised) machine learning experiment to hold out part of the available data as a test set X_test, y_test. The best parameters will be determined by grid search techniques. Below this cell is an attempt to reach a batter MAE with a random forest model; however, this model failed to outperform our ridge regression model, so we will continue optimizing our ridge model with grid search to prevent overfitting. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomForestRegressor(bootstrap=True, criterion='mse', max_depth=None,\n",
       "                      max_features='auto', max_leaf_nodes=None,\n",
       "                      min_impurity_decrease=0.0, min_impurity_split=None,\n",
       "                      min_samples_leaf=1, min_samples_split=2,\n",
       "                      min_weight_fraction_leaf=0.0, n_estimators=10,\n",
       "                      n_jobs=None, oob_score=False, random_state=42, verbose=0,\n",
       "                      warm_start=False)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.ensemble import RandomForestRegressor\n",
    "\n",
    "forest_reg = RandomForestRegressor(n_estimators=10, random_state=42)\n",
    "forest_reg.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2.3858534428413036"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_pred = forest_reg.predict(X_test)\n",
    "forest_mse = mean_squared_error(y_test, y_pred)\n",
    "forest_rmse = np.sqrt(forest_mse)\n",
    "forest_rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scores: [2.39731815 2.09778897 2.26612651 2.17973107 2.24343353 2.36767781\n",
      " 2.17874875 2.50951483 2.16357295 2.36752606]\n",
      "Mean: 2.277143862498279\n",
      "Standard deviation: 0.12259633405870302\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "forest_scores = cross_val_score(forest_reg,  X_train, y_train,\n",
    "                                scoring=\"neg_mean_squared_error\", cv=10)\n",
    "forest_rmse_scores = np.sqrt(-forest_scores)\n",
    "display_scores(forest_rmse_scores)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count    10.000000\n",
       "mean      2.166534\n",
       "std       0.136773\n",
       "min       1.987951\n",
       "25%       2.076302\n",
       "50%       2.183679\n",
       "75%       2.241797\n",
       "max       2.432814\n",
       "dtype: float64"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores = cross_val_score(ridge,  X_train, y_train, scoring=\"neg_mean_squared_error\", cv=10)\n",
    "pd.Series(np.sqrt(-scores)).describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Optimizing our Ridge model with grid search. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### We will add feature selection to our preprocessing pipeline by selecting the best features based on univariate statistical tests. Select K best will work by removing all but the highest k scoring features. We will run this in grid search to return our best hyperparameters. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "({'alpha': 0.01, 'solver': 'auto'}, 0.5399167812051203)"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.feature_selection import SelectKBest\n",
    "                        \n",
    "standardScale = StandardScaler()\n",
    "standardScale.fit_transform(X)\n",
    "\n",
    "selectkBest = SelectKBest()\n",
    "X_new = selectkBest.fit_transform(X, y)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_new, y, test_size = 0.25)\n",
    "\n",
    "X_train, X_test, y_train, y_test = train_test_split(X_new, y, random_state=0, test_size = 0.2)\n",
    "\n",
    "# train across 5 folds, that's a total of (12+6)*5=90 rounds of training \n",
    "\n",
    "param  = {'alpha':[0.01, 0.1, 1,10,100],\n",
    "         'solver' : ['auto', 'svd', 'cholesky', 'lsqr', 'sparse_cg', 'sag', 'saga']}\n",
    "grid_search = GridSearchCV(estimator = Ridge(random_state=10,),\n",
    "param_grid = param,scoring= 'r2' ,cv = 5,  n_jobs = -1)\n",
    "grid_search.fit(X_train, y_train)\n",
    "grid_search.best_params_, grid_search.best_score_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'alpha': 0.01, 'solver': 'auto'}"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Ridge(alpha=0.01, copy_X=True, fit_intercept=True, max_iter=None,\n",
       "      normalize=False, random_state=10, solver='auto', tol=0.001)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "nan {'alpha': 0.01, 'solver': 'auto'}\n",
      "nan {'alpha': 0.01, 'solver': 'svd'}\n",
      "nan {'alpha': 0.01, 'solver': 'cholesky'}\n",
      "nan {'alpha': 0.01, 'solver': 'lsqr'}\n",
      "nan {'alpha': 0.01, 'solver': 'sparse_cg'}\n",
      "nan {'alpha': 0.01, 'solver': 'sag'}\n",
      "nan {'alpha': 0.01, 'solver': 'saga'}\n",
      "nan {'alpha': 0.1, 'solver': 'auto'}\n",
      "nan {'alpha': 0.1, 'solver': 'svd'}\n",
      "nan {'alpha': 0.1, 'solver': 'cholesky'}\n",
      "nan {'alpha': 0.1, 'solver': 'lsqr'}\n",
      "nan {'alpha': 0.1, 'solver': 'sparse_cg'}\n",
      "nan {'alpha': 0.1, 'solver': 'sag'}\n",
      "nan {'alpha': 0.1, 'solver': 'saga'}\n",
      "nan {'alpha': 1, 'solver': 'auto'}\n",
      "nan {'alpha': 1, 'solver': 'svd'}\n",
      "nan {'alpha': 1, 'solver': 'cholesky'}\n",
      "nan {'alpha': 1, 'solver': 'lsqr'}\n",
      "nan {'alpha': 1, 'solver': 'sparse_cg'}\n",
      "nan {'alpha': 1, 'solver': 'sag'}\n",
      "nan {'alpha': 1, 'solver': 'saga'}\n",
      "nan {'alpha': 10, 'solver': 'auto'}\n",
      "nan {'alpha': 10, 'solver': 'svd'}\n",
      "nan {'alpha': 10, 'solver': 'cholesky'}\n",
      "nan {'alpha': 10, 'solver': 'lsqr'}\n",
      "nan {'alpha': 10, 'solver': 'sparse_cg'}\n",
      "nan {'alpha': 10, 'solver': 'sag'}\n",
      "nan {'alpha': 10, 'solver': 'saga'}\n",
      "nan {'alpha': 100, 'solver': 'auto'}\n",
      "nan {'alpha': 100, 'solver': 'svd'}\n",
      "nan {'alpha': 100, 'solver': 'cholesky'}\n",
      "nan {'alpha': 100, 'solver': 'lsqr'}\n",
      "nan {'alpha': 100, 'solver': 'sparse_cg'}\n",
      "nan {'alpha': 100, 'solver': 'sag'}\n",
      "nan {'alpha': 100, 'solver': 'saga'}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/ryanmartin/anaconda3/lib/python3.7/site-packages/ipykernel_launcher.py:3: RuntimeWarning: invalid value encountered in sqrt\n",
      "  This is separate from the ipykernel package so we can avoid doing imports until\n"
     ]
    }
   ],
   "source": [
    "cvres = grid_search.cv_results_\n",
    "for mean_score, params in zip(cvres[\"mean_test_score\"], cvres[\"params\"]):\n",
    "    print(np.sqrt(-mean_score), params)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "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>mean_fit_time</th>\n",
       "      <th>std_fit_time</th>\n",
       "      <th>mean_score_time</th>\n",
       "      <th>std_score_time</th>\n",
       "      <th>param_alpha</th>\n",
       "      <th>param_solver</th>\n",
       "      <th>params</th>\n",
       "      <th>split0_test_score</th>\n",
       "      <th>split1_test_score</th>\n",
       "      <th>split2_test_score</th>\n",
       "      <th>split3_test_score</th>\n",
       "      <th>split4_test_score</th>\n",
       "      <th>mean_test_score</th>\n",
       "      <th>std_test_score</th>\n",
       "      <th>rank_test_score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.004836</td>\n",
       "      <td>0.001864</td>\n",
       "      <td>0.000689</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.01</td>\n",
       "      <td>auto</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'auto'}</td>\n",
       "      <td>0.527018</td>\n",
       "      <td>0.553728</td>\n",
       "      <td>0.552654</td>\n",
       "      <td>0.506101</td>\n",
       "      <td>0.560102</td>\n",
       "      <td>0.539917</td>\n",
       "      <td>0.020347</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.002810</td>\n",
       "      <td>0.001037</td>\n",
       "      <td>0.000604</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.01</td>\n",
       "      <td>svd</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'svd'}</td>\n",
       "      <td>0.527018</td>\n",
       "      <td>0.553728</td>\n",
       "      <td>0.552654</td>\n",
       "      <td>0.506101</td>\n",
       "      <td>0.560102</td>\n",
       "      <td>0.539917</td>\n",
       "      <td>0.020347</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.002499</td>\n",
       "      <td>0.001073</td>\n",
       "      <td>0.001442</td>\n",
       "      <td>0.001680</td>\n",
       "      <td>0.01</td>\n",
       "      <td>cholesky</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'cholesky'}</td>\n",
       "      <td>0.527018</td>\n",
       "      <td>0.553728</td>\n",
       "      <td>0.552654</td>\n",
       "      <td>0.506101</td>\n",
       "      <td>0.560102</td>\n",
       "      <td>0.539917</td>\n",
       "      <td>0.020347</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.004818</td>\n",
       "      <td>0.002341</td>\n",
       "      <td>0.000833</td>\n",
       "      <td>0.000425</td>\n",
       "      <td>0.01</td>\n",
       "      <td>lsqr</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'lsqr'}</td>\n",
       "      <td>0.527776</td>\n",
       "      <td>0.552416</td>\n",
       "      <td>0.551985</td>\n",
       "      <td>0.495347</td>\n",
       "      <td>0.560353</td>\n",
       "      <td>0.537573</td>\n",
       "      <td>0.023771</td>\n",
       "      <td>13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.003160</td>\n",
       "      <td>0.000889</td>\n",
       "      <td>0.000774</td>\n",
       "      <td>0.000327</td>\n",
       "      <td>0.01</td>\n",
       "      <td>sparse_cg</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'sparse_cg'}</td>\n",
       "      <td>0.527777</td>\n",
       "      <td>0.552397</td>\n",
       "      <td>0.551984</td>\n",
       "      <td>0.505553</td>\n",
       "      <td>0.560354</td>\n",
       "      <td>0.539609</td>\n",
       "      <td>0.020231</td>\n",
       "      <td>11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.048677</td>\n",
       "      <td>0.005919</td>\n",
       "      <td>0.001509</td>\n",
       "      <td>0.001012</td>\n",
       "      <td>0.01</td>\n",
       "      <td>sag</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'sag'}</td>\n",
       "      <td>0.526838</td>\n",
       "      <td>0.553841</td>\n",
       "      <td>0.552619</td>\n",
       "      <td>0.506125</td>\n",
       "      <td>0.560049</td>\n",
       "      <td>0.539891</td>\n",
       "      <td>0.020362</td>\n",
       "      <td>4</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.089352</td>\n",
       "      <td>0.008292</td>\n",
       "      <td>0.000850</td>\n",
       "      <td>0.000427</td>\n",
       "      <td>0.01</td>\n",
       "      <td>saga</td>\n",
       "      <td>{'alpha': 0.01, 'solver': 'saga'}</td>\n",
       "      <td>0.526642</td>\n",
       "      <td>0.553956</td>\n",
       "      <td>0.552577</td>\n",
       "      <td>0.506150</td>\n",
       "      <td>0.559985</td>\n",
       "      <td>0.539858</td>\n",
       "      <td>0.020377</td>\n",
       "      <td>5</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.001471</td>\n",
       "      <td>0.000287</td>\n",
       "      <td>0.000592</td>\n",
       "      <td>0.000141</td>\n",
       "      <td>0.1</td>\n",
       "      <td>auto</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'auto'}</td>\n",
       "      <td>0.524729</td>\n",
       "      <td>0.555359</td>\n",
       "      <td>0.551967</td>\n",
       "      <td>0.507982</td>\n",
       "      <td>0.559062</td>\n",
       "      <td>0.539815</td>\n",
       "      <td>0.020002</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.001692</td>\n",
       "      <td>0.000223</td>\n",
       "      <td>0.000758</td>\n",
       "      <td>0.000279</td>\n",
       "      <td>0.1</td>\n",
       "      <td>svd</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'svd'}</td>\n",
       "      <td>0.524729</td>\n",
       "      <td>0.555359</td>\n",
       "      <td>0.551967</td>\n",
       "      <td>0.507982</td>\n",
       "      <td>0.559062</td>\n",
       "      <td>0.539815</td>\n",
       "      <td>0.020002</td>\n",
       "      <td>8</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>0.001478</td>\n",
       "      <td>0.000633</td>\n",
       "      <td>0.000769</td>\n",
       "      <td>0.000661</td>\n",
       "      <td>0.1</td>\n",
       "      <td>cholesky</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'cholesky'}</td>\n",
       "      <td>0.524729</td>\n",
       "      <td>0.555359</td>\n",
       "      <td>0.551967</td>\n",
       "      <td>0.507982</td>\n",
       "      <td>0.559062</td>\n",
       "      <td>0.539815</td>\n",
       "      <td>0.020002</td>\n",
       "      <td>6</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>0.002442</td>\n",
       "      <td>0.000520</td>\n",
       "      <td>0.000855</td>\n",
       "      <td>0.000682</td>\n",
       "      <td>0.1</td>\n",
       "      <td>lsqr</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'lsqr'}</td>\n",
       "      <td>0.522592</td>\n",
       "      <td>0.554302</td>\n",
       "      <td>0.551749</td>\n",
       "      <td>0.500781</td>\n",
       "      <td>0.556256</td>\n",
       "      <td>0.537132</td>\n",
       "      <td>0.021939</td>\n",
       "      <td>14</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.002165</td>\n",
       "      <td>0.000187</td>\n",
       "      <td>0.000658</td>\n",
       "      <td>0.000319</td>\n",
       "      <td>0.1</td>\n",
       "      <td>sparse_cg</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'sparse_cg'}</td>\n",
       "      <td>0.525228</td>\n",
       "      <td>0.554316</td>\n",
       "      <td>0.551519</td>\n",
       "      <td>0.500781</td>\n",
       "      <td>0.559235</td>\n",
       "      <td>0.538212</td>\n",
       "      <td>0.022125</td>\n",
       "      <td>12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.045373</td>\n",
       "      <td>0.020257</td>\n",
       "      <td>0.000670</td>\n",
       "      <td>0.000101</td>\n",
       "      <td>0.1</td>\n",
       "      <td>sag</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'sag'}</td>\n",
       "      <td>0.524565</td>\n",
       "      <td>0.555381</td>\n",
       "      <td>0.551910</td>\n",
       "      <td>0.507959</td>\n",
       "      <td>0.558994</td>\n",
       "      <td>0.539757</td>\n",
       "      <td>0.020018</td>\n",
       "      <td>9</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>0.072731</td>\n",
       "      <td>0.025508</td>\n",
       "      <td>0.001050</td>\n",
       "      <td>0.000959</td>\n",
       "      <td>0.1</td>\n",
       "      <td>saga</td>\n",
       "      <td>{'alpha': 0.1, 'solver': 'saga'}</td>\n",
       "      <td>0.524363</td>\n",
       "      <td>0.555402</td>\n",
       "      <td>0.551834</td>\n",
       "      <td>0.507947</td>\n",
       "      <td>0.558896</td>\n",
       "      <td>0.539684</td>\n",
       "      <td>0.020028</td>\n",
       "      <td>10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>0.001180</td>\n",
       "      <td>0.000074</td>\n",
       "      <td>0.000443</td>\n",
       "      <td>0.000010</td>\n",
       "      <td>1</td>\n",
       "      <td>auto</td>\n",
       "      <td>{'alpha': 1, 'solver': 'auto'}</td>\n",
       "      <td>0.516067</td>\n",
       "      <td>0.555018</td>\n",
       "      <td>0.545022</td>\n",
       "      <td>0.507989</td>\n",
       "      <td>0.551725</td>\n",
       "      <td>0.535158</td>\n",
       "      <td>0.019333</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>0.001687</td>\n",
       "      <td>0.000378</td>\n",
       "      <td>0.000496</td>\n",
       "      <td>0.000037</td>\n",
       "      <td>1</td>\n",
       "      <td>svd</td>\n",
       "      <td>{'alpha': 1, 'solver': 'svd'}</td>\n",
       "      <td>0.516067</td>\n",
       "      <td>0.555018</td>\n",
       "      <td>0.545022</td>\n",
       "      <td>0.507989</td>\n",
       "      <td>0.551725</td>\n",
       "      <td>0.535158</td>\n",
       "      <td>0.019333</td>\n",
       "      <td>17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>0.001376</td>\n",
       "      <td>0.000268</td>\n",
       "      <td>0.000511</td>\n",
       "      <td>0.000063</td>\n",
       "      <td>1</td>\n",
       "      <td>cholesky</td>\n",
       "      <td>{'alpha': 1, 'solver': 'cholesky'}</td>\n",
       "      <td>0.516067</td>\n",
       "      <td>0.555018</td>\n",
       "      <td>0.545022</td>\n",
       "      <td>0.507989</td>\n",
       "      <td>0.551725</td>\n",
       "      <td>0.535158</td>\n",
       "      <td>0.019333</td>\n",
       "      <td>15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.002091</td>\n",
       "      <td>0.000212</td>\n",
       "      <td>0.000477</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>1</td>\n",
       "      <td>lsqr</td>\n",
       "      <td>{'alpha': 1, 'solver': 'lsqr'}</td>\n",
       "      <td>0.512692</td>\n",
       "      <td>0.555013</td>\n",
       "      <td>0.542786</td>\n",
       "      <td>0.507558</td>\n",
       "      <td>0.549750</td>\n",
       "      <td>0.533554</td>\n",
       "      <td>0.019592</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>0.002106</td>\n",
       "      <td>0.000197</td>\n",
       "      <td>0.000474</td>\n",
       "      <td>0.000023</td>\n",
       "      <td>1</td>\n",
       "      <td>sparse_cg</td>\n",
       "      <td>{'alpha': 1, 'solver': 'sparse_cg'}</td>\n",
       "      <td>0.515914</td>\n",
       "      <td>0.555013</td>\n",
       "      <td>0.545128</td>\n",
       "      <td>0.507558</td>\n",
       "      <td>0.551288</td>\n",
       "      <td>0.534974</td>\n",
       "      <td>0.019420</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>0.023746</td>\n",
       "      <td>0.010759</td>\n",
       "      <td>0.000651</td>\n",
       "      <td>0.000078</td>\n",
       "      <td>1</td>\n",
       "      <td>sag</td>\n",
       "      <td>{'alpha': 1, 'solver': 'sag'}</td>\n",
       "      <td>0.516036</td>\n",
       "      <td>0.554958</td>\n",
       "      <td>0.545010</td>\n",
       "      <td>0.507984</td>\n",
       "      <td>0.551703</td>\n",
       "      <td>0.535132</td>\n",
       "      <td>0.019323</td>\n",
       "      <td>18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>0.022246</td>\n",
       "      <td>0.005999</td>\n",
       "      <td>0.000573</td>\n",
       "      <td>0.000073</td>\n",
       "      <td>1</td>\n",
       "      <td>saga</td>\n",
       "      <td>{'alpha': 1, 'solver': 'saga'}</td>\n",
       "      <td>0.515888</td>\n",
       "      <td>0.554946</td>\n",
       "      <td>0.544910</td>\n",
       "      <td>0.507994</td>\n",
       "      <td>0.551602</td>\n",
       "      <td>0.535062</td>\n",
       "      <td>0.019320</td>\n",
       "      <td>19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>0.001149</td>\n",
       "      <td>0.000046</td>\n",
       "      <td>0.000444</td>\n",
       "      <td>0.000020</td>\n",
       "      <td>10</td>\n",
       "      <td>auto</td>\n",
       "      <td>{'alpha': 10, 'solver': 'auto'}</td>\n",
       "      <td>0.493486</td>\n",
       "      <td>0.527115</td>\n",
       "      <td>0.509822</td>\n",
       "      <td>0.476339</td>\n",
       "      <td>0.518971</td>\n",
       "      <td>0.505143</td>\n",
       "      <td>0.018227</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>0.001391</td>\n",
       "      <td>0.000097</td>\n",
       "      <td>0.000451</td>\n",
       "      <td>0.000050</td>\n",
       "      <td>10</td>\n",
       "      <td>svd</td>\n",
       "      <td>{'alpha': 10, 'solver': 'svd'}</td>\n",
       "      <td>0.493486</td>\n",
       "      <td>0.527115</td>\n",
       "      <td>0.509822</td>\n",
       "      <td>0.476339</td>\n",
       "      <td>0.518971</td>\n",
       "      <td>0.505143</td>\n",
       "      <td>0.018227</td>\n",
       "      <td>25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>0.000865</td>\n",
       "      <td>0.000052</td>\n",
       "      <td>0.000362</td>\n",
       "      <td>0.000067</td>\n",
       "      <td>10</td>\n",
       "      <td>cholesky</td>\n",
       "      <td>{'alpha': 10, 'solver': 'cholesky'}</td>\n",
       "      <td>0.493486</td>\n",
       "      <td>0.527115</td>\n",
       "      <td>0.509822</td>\n",
       "      <td>0.476339</td>\n",
       "      <td>0.518971</td>\n",
       "      <td>0.505143</td>\n",
       "      <td>0.018227</td>\n",
       "      <td>23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>0.001283</td>\n",
       "      <td>0.000287</td>\n",
       "      <td>0.000342</td>\n",
       "      <td>0.000076</td>\n",
       "      <td>10</td>\n",
       "      <td>lsqr</td>\n",
       "      <td>{'alpha': 10, 'solver': 'lsqr'}</td>\n",
       "      <td>0.493363</td>\n",
       "      <td>0.526371</td>\n",
       "      <td>0.509505</td>\n",
       "      <td>0.476620</td>\n",
       "      <td>0.518931</td>\n",
       "      <td>0.504955</td>\n",
       "      <td>0.017954</td>\n",
       "      <td>28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.002060</td>\n",
       "      <td>0.000107</td>\n",
       "      <td>0.000548</td>\n",
       "      <td>0.000085</td>\n",
       "      <td>10</td>\n",
       "      <td>sparse_cg</td>\n",
       "      <td>{'alpha': 10, 'solver': 'sparse_cg'}</td>\n",
       "      <td>0.493437</td>\n",
       "      <td>0.527127</td>\n",
       "      <td>0.509816</td>\n",
       "      <td>0.476351</td>\n",
       "      <td>0.518964</td>\n",
       "      <td>0.505135</td>\n",
       "      <td>0.018231</td>\n",
       "      <td>27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>0.008981</td>\n",
       "      <td>0.002608</td>\n",
       "      <td>0.001109</td>\n",
       "      <td>0.000854</td>\n",
       "      <td>10</td>\n",
       "      <td>sag</td>\n",
       "      <td>{'alpha': 10, 'solver': 'sag'}</td>\n",
       "      <td>0.493399</td>\n",
       "      <td>0.527137</td>\n",
       "      <td>0.509812</td>\n",
       "      <td>0.476373</td>\n",
       "      <td>0.518980</td>\n",
       "      <td>0.505136</td>\n",
       "      <td>0.018234</td>\n",
       "      <td>26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.019891</td>\n",
       "      <td>0.021393</td>\n",
       "      <td>0.000730</td>\n",
       "      <td>0.000106</td>\n",
       "      <td>10</td>\n",
       "      <td>saga</td>\n",
       "      <td>{'alpha': 10, 'solver': 'saga'}</td>\n",
       "      <td>0.493489</td>\n",
       "      <td>0.527125</td>\n",
       "      <td>0.509835</td>\n",
       "      <td>0.476332</td>\n",
       "      <td>0.518959</td>\n",
       "      <td>0.505144</td>\n",
       "      <td>0.018230</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>0.002037</td>\n",
       "      <td>0.000459</td>\n",
       "      <td>0.000885</td>\n",
       "      <td>0.000215</td>\n",
       "      <td>100</td>\n",
       "      <td>auto</td>\n",
       "      <td>{'alpha': 100, 'solver': 'auto'}</td>\n",
       "      <td>0.397369</td>\n",
       "      <td>0.416622</td>\n",
       "      <td>0.397240</td>\n",
       "      <td>0.375514</td>\n",
       "      <td>0.401856</td>\n",
       "      <td>0.397720</td>\n",
       "      <td>0.013173</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>0.002418</td>\n",
       "      <td>0.001245</td>\n",
       "      <td>0.001525</td>\n",
       "      <td>0.001512</td>\n",
       "      <td>100</td>\n",
       "      <td>svd</td>\n",
       "      <td>{'alpha': 100, 'solver': 'svd'}</td>\n",
       "      <td>0.397369</td>\n",
       "      <td>0.416622</td>\n",
       "      <td>0.397240</td>\n",
       "      <td>0.375514</td>\n",
       "      <td>0.401856</td>\n",
       "      <td>0.397720</td>\n",
       "      <td>0.013173</td>\n",
       "      <td>30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>0.001573</td>\n",
       "      <td>0.000490</td>\n",
       "      <td>0.000718</td>\n",
       "      <td>0.000351</td>\n",
       "      <td>100</td>\n",
       "      <td>cholesky</td>\n",
       "      <td>{'alpha': 100, 'solver': 'cholesky'}</td>\n",
       "      <td>0.397369</td>\n",
       "      <td>0.416622</td>\n",
       "      <td>0.397240</td>\n",
       "      <td>0.375514</td>\n",
       "      <td>0.401856</td>\n",
       "      <td>0.397720</td>\n",
       "      <td>0.013173</td>\n",
       "      <td>31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.002057</td>\n",
       "      <td>0.000663</td>\n",
       "      <td>0.000451</td>\n",
       "      <td>0.000029</td>\n",
       "      <td>100</td>\n",
       "      <td>lsqr</td>\n",
       "      <td>{'alpha': 100, 'solver': 'lsqr'}</td>\n",
       "      <td>0.397370</td>\n",
       "      <td>0.416619</td>\n",
       "      <td>0.397239</td>\n",
       "      <td>0.375392</td>\n",
       "      <td>0.401856</td>\n",
       "      <td>0.397695</td>\n",
       "      <td>0.013213</td>\n",
       "      <td>35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>0.002000</td>\n",
       "      <td>0.000363</td>\n",
       "      <td>0.000453</td>\n",
       "      <td>0.000032</td>\n",
       "      <td>100</td>\n",
       "      <td>sparse_cg</td>\n",
       "      <td>{'alpha': 100, 'solver': 'sparse_cg'}</td>\n",
       "      <td>0.397370</td>\n",
       "      <td>0.416619</td>\n",
       "      <td>0.397239</td>\n",
       "      <td>0.375515</td>\n",
       "      <td>0.401856</td>\n",
       "      <td>0.397720</td>\n",
       "      <td>0.013171</td>\n",
       "      <td>33</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>0.004421</td>\n",
       "      <td>0.000933</td>\n",
       "      <td>0.000438</td>\n",
       "      <td>0.000098</td>\n",
       "      <td>100</td>\n",
       "      <td>sag</td>\n",
       "      <td>{'alpha': 100, 'solver': 'sag'}</td>\n",
       "      <td>0.397369</td>\n",
       "      <td>0.416635</td>\n",
       "      <td>0.397224</td>\n",
       "      <td>0.375503</td>\n",
       "      <td>0.401846</td>\n",
       "      <td>0.397715</td>\n",
       "      <td>0.013179</td>\n",
       "      <td>34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>0.005365</td>\n",
       "      <td>0.001204</td>\n",
       "      <td>0.000473</td>\n",
       "      <td>0.000091</td>\n",
       "      <td>100</td>\n",
       "      <td>saga</td>\n",
       "      <td>{'alpha': 100, 'solver': 'saga'}</td>\n",
       "      <td>0.397355</td>\n",
       "      <td>0.416676</td>\n",
       "      <td>0.397261</td>\n",
       "      <td>0.375501</td>\n",
       "      <td>0.401846</td>\n",
       "      <td>0.397728</td>\n",
       "      <td>0.013192</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    mean_fit_time  std_fit_time  mean_score_time  std_score_time param_alpha  \\\n",
       "0        0.004836      0.001864         0.000689        0.000086        0.01   \n",
       "1        0.002810      0.001037         0.000604        0.000038        0.01   \n",
       "2        0.002499      0.001073         0.001442        0.001680        0.01   \n",
       "3        0.004818      0.002341         0.000833        0.000425        0.01   \n",
       "4        0.003160      0.000889         0.000774        0.000327        0.01   \n",
       "5        0.048677      0.005919         0.001509        0.001012        0.01   \n",
       "6        0.089352      0.008292         0.000850        0.000427        0.01   \n",
       "7        0.001471      0.000287         0.000592        0.000141         0.1   \n",
       "8        0.001692      0.000223         0.000758        0.000279         0.1   \n",
       "9        0.001478      0.000633         0.000769        0.000661         0.1   \n",
       "10       0.002442      0.000520         0.000855        0.000682         0.1   \n",
       "11       0.002165      0.000187         0.000658        0.000319         0.1   \n",
       "12       0.045373      0.020257         0.000670        0.000101         0.1   \n",
       "13       0.072731      0.025508         0.001050        0.000959         0.1   \n",
       "14       0.001180      0.000074         0.000443        0.000010           1   \n",
       "15       0.001687      0.000378         0.000496        0.000037           1   \n",
       "16       0.001376      0.000268         0.000511        0.000063           1   \n",
       "17       0.002091      0.000212         0.000477        0.000038           1   \n",
       "18       0.002106      0.000197         0.000474        0.000023           1   \n",
       "19       0.023746      0.010759         0.000651        0.000078           1   \n",
       "20       0.022246      0.005999         0.000573        0.000073           1   \n",
       "21       0.001149      0.000046         0.000444        0.000020          10   \n",
       "22       0.001391      0.000097         0.000451        0.000050          10   \n",
       "23       0.000865      0.000052         0.000362        0.000067          10   \n",
       "24       0.001283      0.000287         0.000342        0.000076          10   \n",
       "25       0.002060      0.000107         0.000548        0.000085          10   \n",
       "26       0.008981      0.002608         0.001109        0.000854          10   \n",
       "27       0.019891      0.021393         0.000730        0.000106          10   \n",
       "28       0.002037      0.000459         0.000885        0.000215         100   \n",
       "29       0.002418      0.001245         0.001525        0.001512         100   \n",
       "30       0.001573      0.000490         0.000718        0.000351         100   \n",
       "31       0.002057      0.000663         0.000451        0.000029         100   \n",
       "32       0.002000      0.000363         0.000453        0.000032         100   \n",
       "33       0.004421      0.000933         0.000438        0.000098         100   \n",
       "34       0.005365      0.001204         0.000473        0.000091         100   \n",
       "\n",
       "   param_solver                                  params  split0_test_score  \\\n",
       "0          auto       {'alpha': 0.01, 'solver': 'auto'}           0.527018   \n",
       "1           svd        {'alpha': 0.01, 'solver': 'svd'}           0.527018   \n",
       "2      cholesky   {'alpha': 0.01, 'solver': 'cholesky'}           0.527018   \n",
       "3          lsqr       {'alpha': 0.01, 'solver': 'lsqr'}           0.527776   \n",
       "4     sparse_cg  {'alpha': 0.01, 'solver': 'sparse_cg'}           0.527777   \n",
       "5           sag        {'alpha': 0.01, 'solver': 'sag'}           0.526838   \n",
       "6          saga       {'alpha': 0.01, 'solver': 'saga'}           0.526642   \n",
       "7          auto        {'alpha': 0.1, 'solver': 'auto'}           0.524729   \n",
       "8           svd         {'alpha': 0.1, 'solver': 'svd'}           0.524729   \n",
       "9      cholesky    {'alpha': 0.1, 'solver': 'cholesky'}           0.524729   \n",
       "10         lsqr        {'alpha': 0.1, 'solver': 'lsqr'}           0.522592   \n",
       "11    sparse_cg   {'alpha': 0.1, 'solver': 'sparse_cg'}           0.525228   \n",
       "12          sag         {'alpha': 0.1, 'solver': 'sag'}           0.524565   \n",
       "13         saga        {'alpha': 0.1, 'solver': 'saga'}           0.524363   \n",
       "14         auto          {'alpha': 1, 'solver': 'auto'}           0.516067   \n",
       "15          svd           {'alpha': 1, 'solver': 'svd'}           0.516067   \n",
       "16     cholesky      {'alpha': 1, 'solver': 'cholesky'}           0.516067   \n",
       "17         lsqr          {'alpha': 1, 'solver': 'lsqr'}           0.512692   \n",
       "18    sparse_cg     {'alpha': 1, 'solver': 'sparse_cg'}           0.515914   \n",
       "19          sag           {'alpha': 1, 'solver': 'sag'}           0.516036   \n",
       "20         saga          {'alpha': 1, 'solver': 'saga'}           0.515888   \n",
       "21         auto         {'alpha': 10, 'solver': 'auto'}           0.493486   \n",
       "22          svd          {'alpha': 10, 'solver': 'svd'}           0.493486   \n",
       "23     cholesky     {'alpha': 10, 'solver': 'cholesky'}           0.493486   \n",
       "24         lsqr         {'alpha': 10, 'solver': 'lsqr'}           0.493363   \n",
       "25    sparse_cg    {'alpha': 10, 'solver': 'sparse_cg'}           0.493437   \n",
       "26          sag          {'alpha': 10, 'solver': 'sag'}           0.493399   \n",
       "27         saga         {'alpha': 10, 'solver': 'saga'}           0.493489   \n",
       "28         auto        {'alpha': 100, 'solver': 'auto'}           0.397369   \n",
       "29          svd         {'alpha': 100, 'solver': 'svd'}           0.397369   \n",
       "30     cholesky    {'alpha': 100, 'solver': 'cholesky'}           0.397369   \n",
       "31         lsqr        {'alpha': 100, 'solver': 'lsqr'}           0.397370   \n",
       "32    sparse_cg   {'alpha': 100, 'solver': 'sparse_cg'}           0.397370   \n",
       "33          sag         {'alpha': 100, 'solver': 'sag'}           0.397369   \n",
       "34         saga        {'alpha': 100, 'solver': 'saga'}           0.397355   \n",
       "\n",
       "    split1_test_score  split2_test_score  split3_test_score  \\\n",
       "0            0.553728           0.552654           0.506101   \n",
       "1            0.553728           0.552654           0.506101   \n",
       "2            0.553728           0.552654           0.506101   \n",
       "3            0.552416           0.551985           0.495347   \n",
       "4            0.552397           0.551984           0.505553   \n",
       "5            0.553841           0.552619           0.506125   \n",
       "6            0.553956           0.552577           0.506150   \n",
       "7            0.555359           0.551967           0.507982   \n",
       "8            0.555359           0.551967           0.507982   \n",
       "9            0.555359           0.551967           0.507982   \n",
       "10           0.554302           0.551749           0.500781   \n",
       "11           0.554316           0.551519           0.500781   \n",
       "12           0.555381           0.551910           0.507959   \n",
       "13           0.555402           0.551834           0.507947   \n",
       "14           0.555018           0.545022           0.507989   \n",
       "15           0.555018           0.545022           0.507989   \n",
       "16           0.555018           0.545022           0.507989   \n",
       "17           0.555013           0.542786           0.507558   \n",
       "18           0.555013           0.545128           0.507558   \n",
       "19           0.554958           0.545010           0.507984   \n",
       "20           0.554946           0.544910           0.507994   \n",
       "21           0.527115           0.509822           0.476339   \n",
       "22           0.527115           0.509822           0.476339   \n",
       "23           0.527115           0.509822           0.476339   \n",
       "24           0.526371           0.509505           0.476620   \n",
       "25           0.527127           0.509816           0.476351   \n",
       "26           0.527137           0.509812           0.476373   \n",
       "27           0.527125           0.509835           0.476332   \n",
       "28           0.416622           0.397240           0.375514   \n",
       "29           0.416622           0.397240           0.375514   \n",
       "30           0.416622           0.397240           0.375514   \n",
       "31           0.416619           0.397239           0.375392   \n",
       "32           0.416619           0.397239           0.375515   \n",
       "33           0.416635           0.397224           0.375503   \n",
       "34           0.416676           0.397261           0.375501   \n",
       "\n",
       "    split4_test_score  mean_test_score  std_test_score  rank_test_score  \n",
       "0            0.560102         0.539917        0.020347                1  \n",
       "1            0.560102         0.539917        0.020347                3  \n",
       "2            0.560102         0.539917        0.020347                1  \n",
       "3            0.560353         0.537573        0.023771               13  \n",
       "4            0.560354         0.539609        0.020231               11  \n",
       "5            0.560049         0.539891        0.020362                4  \n",
       "6            0.559985         0.539858        0.020377                5  \n",
       "7            0.559062         0.539815        0.020002                6  \n",
       "8            0.559062         0.539815        0.020002                8  \n",
       "9            0.559062         0.539815        0.020002                6  \n",
       "10           0.556256         0.537132        0.021939               14  \n",
       "11           0.559235         0.538212        0.022125               12  \n",
       "12           0.558994         0.539757        0.020018                9  \n",
       "13           0.558896         0.539684        0.020028               10  \n",
       "14           0.551725         0.535158        0.019333               15  \n",
       "15           0.551725         0.535158        0.019333               17  \n",
       "16           0.551725         0.535158        0.019333               15  \n",
       "17           0.549750         0.533554        0.019592               21  \n",
       "18           0.551288         0.534974        0.019420               20  \n",
       "19           0.551703         0.535132        0.019323               18  \n",
       "20           0.551602         0.535062        0.019320               19  \n",
       "21           0.518971         0.505143        0.018227               23  \n",
       "22           0.518971         0.505143        0.018227               25  \n",
       "23           0.518971         0.505143        0.018227               23  \n",
       "24           0.518931         0.504955        0.017954               28  \n",
       "25           0.518964         0.505135        0.018231               27  \n",
       "26           0.518980         0.505136        0.018234               26  \n",
       "27           0.518959         0.505144        0.018230               22  \n",
       "28           0.401856         0.397720        0.013173               31  \n",
       "29           0.401856         0.397720        0.013173               30  \n",
       "30           0.401856         0.397720        0.013173               31  \n",
       "31           0.401856         0.397695        0.013213               35  \n",
       "32           0.401856         0.397720        0.013171               33  \n",
       "33           0.401846         0.397715        0.013179               34  \n",
       "34           0.401846         0.397728        0.013192               29  "
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pd.DataFrame(grid_search.cv_results_)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluate our best model on the test set"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Below are the results of our final model using ridge regression and grid search cv to find the best hyperparameters. This model will basically predict the age of the abalone within 2.1 years of its actual age (see final MAE) without overfitting. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_model = grid_search.best_estimator_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_pred = final_model.predict(X_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0, 0.5, 'Predicted Y')"
      ]
     },
     "execution_count": 65,
     "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": [
    "plt.scatter(y_test, y_pred)\n",
    "plt.xlabel('Y Test')\n",
    "plt.ylabel('Predicted Y')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "final_mse = mean_squared_error(y_test, y_pred)\n",
    "final_rmse = np.sqrt(final_mse)\n",
    "final_mae = mean_absolute_error(y_test, y_pred)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('MSE:', 4.832429123001742)"
      ]
     },
     "execution_count": 67,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"MSE:\", final_mse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('RMSE:', 2.198278672735043)"
      ]
     },
     "execution_count": 68,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"RMSE:\", final_rmse"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('RMSE:', 1.5469163072850936)"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "\"RMSE:\", final_mae"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Scores: [2.25355172 1.99070742 2.07545853 2.2255554  1.98744618 2.21091052\n",
      " 2.0847361  2.43249349 2.15632327 2.24720639]\n",
      "Mean: 2.1664389006674605\n",
      "Standard deviation: 0.1297476823814643\n"
     ]
    }
   ],
   "source": [
    "from sklearn.model_selection import cross_val_score\n",
    "\n",
    "final_scores = cross_val_score(final_model,  X_train, y_train,\n",
    "                                scoring=\"neg_mean_squared_error\", cv=10)\n",
    "final_rmse_scores = np.sqrt(-final_scores)\n",
    "display_scores(final_rmse_scores)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Residual Plot:\n",
    "\n",
    "In regression analysis, the difference between the observed value of the dependent variable (y) and the predicted value (ŷ) is called the residual (e). Each data point has one residual.\n",
    "\n",
    "Residual = Observed value - Predicted value\n",
    "e = y - ŷ\n",
    "\n",
    "Both the sum and the mean of the residuals are equal to zero. That is, Σ e = 0 and e = 0."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.hist((y_test-y_pred),bins=50);"
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}