vicperotti/Wine+Reviews+from+Kaggle-With+FeatureNames.ipynb

## Wine+Reviews+from+Kaggle-With+FeatureNames.ipynb
{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# In this example, Let's look at preprocessing in sklearn\n",
    "specifically I want to highlight....\n",
    "encoding categorical variables\n",
    "and imputing missing values\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# This example from https://www.kaggle.com/zynicide/wine-reviews\n",
    "# 2. Import libraries and modules\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import joblib\n",
    "\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "from sklearn.ensemble import RandomForestRegressor\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.model_selection import GridSearchCV\n",
    "from sklearn.metrics import mean_squared_error, r2_score\n",
    "from sklearn.compose import ColumnTransformer, make_column_transformer\n",
    "\n",
    "import multiprocessing\n",
    "from afinn import Afinn"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "8"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "cores = multiprocessing.cpu_count()\n",
    "cores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "129971"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = pd.read_csv('winemag-data-130k-v2.csv') \n",
    "#data = pd.read_csv('winemag-data_first150k.csv')\n",
    "afinn = Afinn(emoticons=True)\n",
    "len(data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Unnamed: 0                   0\n",
      "country                     63\n",
      "description                  0\n",
      "designation              37465\n",
      "points                       0\n",
      "price                     8996\n",
      "province                    63\n",
      "region_1                 21247\n",
      "region_2                 79460\n",
      "taster_name              26244\n",
      "taster_twitter_handle    31213\n",
      "title                        0\n",
      "variety                      1\n",
      "winery                       0\n",
      "dtype: int64\n"
     ]
    }
   ],
   "source": [
    "# look a number of columns have missing values!\n",
    "print(data.isnull().sum())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(54971, 14)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#drop a bunch of rows to make this more quick to compute\n",
    "np.random.seed(10)\n",
    "remove_n = 75000\n",
    "drop_indices = np.random.choice(data.index, remove_n, replace=False)\n",
    "data = data.drop(drop_indices)\n",
    "data.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "after dropping na there are 38602 rows left\n",
      "   country  sentiment  price    province taster_twitter_handle  \\\n",
      "2       US        0.0   14.0      Oregon           @paulgwine    \n",
      "4       US        5.0   65.0      Oregon           @paulgwine    \n",
      "10      US        6.0   19.0  California               @vboone   \n",
      "12      US        2.0   34.0  California               @vboone   \n",
      "21      US        2.0   20.0      Oregon           @paulgwine    \n",
      "\n",
      "               variety              winery  \n",
      "2           Pinot Gris           Rainstorm  \n",
      "4           Pinot Noir        Sweet Cheeks  \n",
      "10  Cabernet Sauvignon  Kirkland Signature  \n",
      "12  Cabernet Sauvignon    Louis M. Martini  \n",
      "21          Pinot Noir             Acrobat  \n"
     ]
    }
   ],
   "source": [
    "#Oh why not add sentiment to description\n",
    "data[['sentiment']] = data['description'].apply(lambda Text: pd.Series(afinn.score(Text)))\n",
    "\n",
    "# DROP rows with missing values in each of the following\n",
    "data2 = data.dropna(subset=['points','country','sentiment','price', 'province','taster_twitter_handle','variety','winery'])\n",
    "features = data2[['country','sentiment','price', 'province','taster_twitter_handle','variety','winery']]\n",
    "print(\"after dropping na there are %d rows left\" % len(data2))\n",
    "print(features.head())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "y = data2.points\n",
    "#X = data.drop('points', axis=1)\n",
    "X_train, X_test, y_train, y_test = train_test_split(features, y,\n",
    "                                                    test_size=0.33,\n",
    "                                                    random_state=42)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# so how do we handle those pesky categorical variables\n",
    "#VIA https://scikit-learn.org/stable/modules/preprocessing.html\n",
    "#AND https://jorisvandenbossche.github.io/blog/2018/05/28/scikit-learn-columntransformer/\n",
    "\n",
    "preprocess = make_column_transformer(\n",
    "    (preprocessing.OneHotEncoder(handle_unknown='ignore'),['country', 'province', 'taster_twitter_handle','variety','winery'] ),\n",
    "    (preprocessing.StandardScaler(),['price', 'sentiment'])\n",
    ")\n",
    "\n",
    "# 5. Declare data preprocessing steps\n",
    "pipeline = make_pipeline(preprocess,\n",
    "                         RandomForestRegressor(n_estimators=100,n_jobs=4))\n",
    "\n",
    "# 6. Declare hyperparameters to tune\n",
    "hyperparameters = { 'randomforestregressor__max_features' : ['auto', 'sqrt', 'log2'],\n",
    "                  'randomforestregressor__max_depth': [None, 5, 3, 1]}\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 loop, best of 3: 6min 5s per loop\n"
     ]
    }
   ],
   "source": [
    "# 7. Tune model using cross-validation pipeline\n",
    "# this is grid search with dropped na values\n",
    "\n",
    "clf = GridSearchCV(pipeline, hyperparameters, cv=5,n_jobs=4)\n",
    "\n",
    "%timeit clf.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5742140205014439\n",
      "3.7222053748721757\n"
     ]
    }
   ],
   "source": [
    "# 9. Evaluate model pipeline on test data\n",
    "pred = clf.predict(X_test)\n",
    "print (r2_score(y_test, pred))\n",
    "print (mean_squared_error(y_test, pred))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": false,
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=5, error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('columntransformer',\n",
       "                                        ColumnTransformer(n_jobs=None,\n",
       "                                                          remainder='drop',\n",
       "                                                          sparse_threshold=0.3,\n",
       "                                                          transformer_weights=None,\n",
       "                                                          transformers=[('onehotencoder',\n",
       "                                                                         OneHotEncoder(categories='auto',\n",
       "                                                                                       drop=None,\n",
       "                                                                                       dtype=<class 'numpy.float64'>,\n",
       "                                                                                       handle_unknown='ignore',\n",
       "                                                                                       sparse=True),\n",
       "                                                                         ['country',\n",
       "                                                                          'province',\n",
       "                                                                          'ta...\n",
       "                                                              min_weight_fraction_leaf=0.0,\n",
       "                                                              n_estimators=100,\n",
       "                                                              n_jobs=4,\n",
       "                                                              oob_score=False,\n",
       "                                                              random_state=None,\n",
       "                                                              verbose=0,\n",
       "                                                              warm_start=False))],\n",
       "                                verbose=False),\n",
       "             iid='deprecated', n_jobs=4,\n",
       "             param_grid={'randomforestregressor__max_depth': [None, 5, 3, 1],\n",
       "                         'randomforestregressor__max_features': ['auto', 'sqrt',\n",
       "                                                                 'log2']},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(9372,)"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "### imported may have to fix steps\n",
    "imports = clf.best_estimator_.steps[1][1].feature_importances_\n",
    "imports.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "import warnings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "#via https://johaupt.github.io/scikit-learn/tutorial/python/data%20processing/ml%20pipeline/model%20interpretation/columnTransformer_feature_names.html\n",
    "def get_feature_names(column_transformer):\n",
    "    \"\"\"Get feature names from all transformers.\n",
    "    Returns\n",
    "    -------\n",
    "    feature_names : list of strings\n",
    "        Names of the features produced by transform.\n",
    "    \"\"\"\n",
    "    # Remove the internal helper function\n",
    "    #check_is_fitted(column_transformer)\n",
    "    \n",
    "    # Turn loopkup into function for better handling with pipeline later\n",
    "    def get_names(trans):\n",
    "        # >> Original get_feature_names() method\n",
    "        if trans == 'drop' or (\n",
    "                hasattr(column, '__len__') and not len(column)):\n",
    "            return []\n",
    "        if trans == 'passthrough':\n",
    "            if hasattr(column_transformer, '_df_columns'):\n",
    "                if ((not isinstance(column, slice))\n",
    "                        and all(isinstance(col, str) for col in column)):\n",
    "                    return column\n",
    "                else:\n",
    "                    return column_transformer._df_columns[column]\n",
    "            else:\n",
    "                indices = np.arange(column_transformer._n_features)\n",
    "                return ['x%d' % i for i in indices[column]]\n",
    "        if not hasattr(trans, 'get_feature_names'):\n",
    "        # >>> Change: Return input column names if no method avaiable\n",
    "            # Turn error into a warning\n",
    "            warnings.warn(\"Transformer %s (type %s) does not \"\n",
    "                                 \"provide get_feature_names. \"\n",
    "                                 \"Will return input column names if available\"\n",
    "                                 % (str(name), type(trans).__name__))\n",
    "            # For transformers without a get_features_names method, use the input\n",
    "            # names to the column transformer\n",
    "            if column is None:\n",
    "                return []\n",
    "            else:\n",
    "                return [name + \"__\" + f for f in column]\n",
    "\n",
    "        return [name + \"__\" + f for f in trans.get_feature_names()]\n",
    "    \n",
    "    ### Start of processing\n",
    "    feature_names = []\n",
    "    \n",
    "    # Allow transformers to be pipelines. Pipeline steps are named differently, so preprocessing is needed\n",
    "    if type(column_transformer) == Pipeline: #sklearn.pipeline.Pipeline:\n",
    "        l_transformers = [(name, trans, None, None) for step, name, trans in column_transformer._iter()]\n",
    "    else:\n",
    "        # For column transformers, follow the original method\n",
    "        l_transformers = list(column_transformer._iter(fitted=True))\n",
    "    \n",
    "    \n",
    "    for name, trans, column, _ in l_transformers: \n",
    "        if type(trans) == Pipeline: #sklearn.pipeline.Pipeline:\n",
    "            # Recursive call on pipeline\n",
    "            _names = get_feature_names(trans)\n",
    "            # if pipeline has no transformer that returns names\n",
    "            if len(_names)==0:\n",
    "                _names = [name + \"__\" + f for f in column]\n",
    "            feature_names.extend(_names)\n",
    "        else:\n",
    "            feature_names.extend(get_names(trans))\n",
    "    \n",
    "    return feature_names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\vjpbbu\\appdata\\local\\programs\\python\\python35-32\\lib\\site-packages\\ipykernel\\__main__.py:33: UserWarning: Transformer standardscaler (type StandardScaler) does not provide get_feature_names. Will return input column names if available\n"
     ]
    }
   ],
   "source": [
    "featnames = get_feature_names(clf.best_estimator_.steps[0][1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9372"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(featnames)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x1d3913f0>"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAz0AAAFdCAYAAAApAwMGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x1d391d70>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "feat_importances = pd.Series(imports, index=featnames)\n",
    "feat_importances.nlargest(20).plot(kind='barh')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['wineregressor.pkl']"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 10. Save model for future use\n",
    "joblib.dump(clf, 'wineregressor.pkl')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Impute missing values"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# what would happen if we add imputer to fill in missing values rather than drop them?\n",
    "# We create the preprocessing pipelines for both numeric and categorical data.\n",
    "# via https://scikit-learn.org/stable/auto_examples/compose/plot_column_transformer_mixed_types.html\n",
    "\n",
    "from sklearn.impute import SimpleImputer\n",
    "from sklearn.pipeline import Pipeline\n",
    "numeric_features = ['price', 'sentiment']\n",
    "numeric_transformer = Pipeline(steps=[\n",
    "    ('imputer', SimpleImputer(strategy='median')),\n",
    "    ('scaler', preprocessing.StandardScaler())])\n",
    "\n",
    "categorical_features = ['country', 'province', 'taster_twitter_handle','variety','winery']\n",
    "categorical_transformer = Pipeline(steps=[\n",
    "    ('imputer', SimpleImputer(strategy='constant', fill_value='missing')),\n",
    "    ('onehot', preprocessing.OneHotEncoder(handle_unknown='ignore'))])\n",
    "\n",
    "preprocessor = ColumnTransformer(\n",
    "    transformers=[\n",
    "        ('num', numeric_transformer, numeric_features),\n",
    "        ('cat', categorical_transformer, categorical_features)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# Append classifier to preprocessing pipeline.\n",
    "# Now we have a full prediction pipeline.\n",
    "clf2 = Pipeline(steps=[('preprocessor', preprocessor),\n",
    "                      ('rfregressor', RandomForestRegressor(n_estimators=100))])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 ColumnTransformer(n_jobs=None, remainder='drop',\n",
       "                                   sparse_threshold=0.3,\n",
       "                                   transformer_weights=None,\n",
       "                                   transformers=[('num',\n",
       "                                                  Pipeline(memory=None,\n",
       "                                                           steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=False,\n",
       "                                                                                 copy=True,\n",
       "                                                                                 fill_value=None,\n",
       "                                                                                 missing_values=nan,\n",
       "                                                                                 strategy='median',\n",
       "                                                                                 verbose=0)),\n",
       "                                                                  ('scaler',\n",
       "                                                                   StandardScaler(copy=True,\n",
       "                                                                                  with_mean...\n",
       "                 RandomForestRegressor(bootstrap=True, ccp_alpha=0.0,\n",
       "                                       criterion='mse', max_depth=None,\n",
       "                                       max_features='auto', max_leaf_nodes=None,\n",
       "                                       max_samples=None,\n",
       "                                       min_impurity_decrease=0.0,\n",
       "                                       min_impurity_split=None,\n",
       "                                       min_samples_leaf=1, min_samples_split=2,\n",
       "                                       min_weight_fraction_leaf=0.0,\n",
       "                                       n_estimators=100, n_jobs=None,\n",
       "                                       oob_score=False, random_state=None,\n",
       "                                       verbose=0, warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 7. Tune model using cross-validation pipeline\n",
    "\n",
    "clf2.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5412127084613201\n",
      "4.010701631132768\n"
     ]
    }
   ],
   "source": [
    "# 9. Evaluate model pipeline on test data\n",
    "pred2 = clf2.predict(X_test)\n",
    "print (r2_score(y_test, pred2))\n",
    "print (mean_squared_error(y_test, pred2))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Pipeline(memory=None,\n",
       "         steps=[('preprocessor',\n",
       "                 ColumnTransformer(n_jobs=None, remainder='drop',\n",
       "                                   sparse_threshold=0.3,\n",
       "                                   transformer_weights=None,\n",
       "                                   transformers=[('num',\n",
       "                                                  Pipeline(memory=None,\n",
       "                                                           steps=[('imputer',\n",
       "                                                                   SimpleImputer(add_indicator=False,\n",
       "                                                                                 copy=True,\n",
       "                                                                                 fill_value=None,\n",
       "                                                                                 missing_values=nan,\n",
       "                                                                                 strategy='median',\n",
       "                                                                                 verbose=0)),\n",
       "                                                                  ('scaler',\n",
       "                                                                   StandardScaler(copy=True,\n",
       "                                                                                  with_mean...\n",
       "                 RandomForestRegressor(bootstrap=True, ccp_alpha=0.0,\n",
       "                                       criterion='mse', max_depth=None,\n",
       "                                       max_features='auto', max_leaf_nodes=None,\n",
       "                                       max_samples=None,\n",
       "                                       min_impurity_decrease=0.0,\n",
       "                                       min_impurity_split=None,\n",
       "                                       min_samples_leaf=1, min_samples_split=2,\n",
       "                                       min_weight_fraction_leaf=0.0,\n",
       "                                       n_estimators=100, n_jobs=None,\n",
       "                                       oob_score=False, random_state=None,\n",
       "                                       verbose=0, warm_start=False))],\n",
       "         verbose=False)"
      ]
     },
     "execution_count": 52,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([4.24227524e-01, 9.54154843e-02, 8.73052284e-04, ...,\n",
       "       3.89830918e-06, 9.46673014e-07, 1.38878325e-04])"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clf2.steps[1][1].feature_importances_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\vjpbbu\\appdata\\local\\programs\\python\\python35-32\\lib\\site-packages\\ipykernel\\__main__.py:33: UserWarning: Transformer imputer (type SimpleImputer) does not provide get_feature_names. Will return input column names if available\n",
      "c:\\users\\vjpbbu\\appdata\\local\\programs\\python\\python35-32\\lib\\site-packages\\ipykernel\\__main__.py:33: UserWarning: Transformer scaler (type StandardScaler) does not provide get_feature_names. Will return input column names if available\n"
     ]
    }
   ],
   "source": [
    "featimports2 = clf2.steps[1][1].feature_importances_\n",
    "featnames2 = get_feature_names(clf2.steps[0][1])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x2096caf0>"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAt4AAAFdCAYAAAA0WGq0AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nH24iaMB7RmxPAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x204bc970>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "%matplotlib inline\n",
    "feat_importances2 = pd.Series(featimports2, index=featnames2)\n",
    "feat_importances2.nlargest(20).plot(kind='barh')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "pipeline2 = make_pipeline(preprocessor,RandomForestRegressor(n_estimators=100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "GridSearchCV(cv=10, error_score=nan,\n",
       "             estimator=Pipeline(memory=None,\n",
       "                                steps=[('columntransformer',\n",
       "                                        ColumnTransformer(n_jobs=None,\n",
       "                                                          remainder='drop',\n",
       "                                                          sparse_threshold=0.3,\n",
       "                                                          transformer_weights=None,\n",
       "                                                          transformers=[('num',\n",
       "                                                                         Pipeline(memory=None,\n",
       "                                                                                  steps=[('imputer',\n",
       "                                                                                          SimpleImputer(add_indicator=False,\n",
       "                                                                                                        copy=True,\n",
       "                                                                                                        fill_value=None,\n",
       "                                                                                                        missing_values=nan,\n",
       "                                                                                                        strategy='median',\n",
       "                                                                                                        verbos...\n",
       "                                                              min_weight_fraction_leaf=0.0,\n",
       "                                                              n_estimators=100,\n",
       "                                                              n_jobs=None,\n",
       "                                                              oob_score=False,\n",
       "                                                              random_state=None,\n",
       "                                                              verbose=0,\n",
       "                                                              warm_start=False))],\n",
       "                                verbose=False),\n",
       "             iid='deprecated', n_jobs=None,\n",
       "             param_grid={'randomforestregressor__max_depth': [None, 5, 3, 1],\n",
       "                         'randomforestregressor__max_features': ['auto', 'sqrt',\n",
       "                                                                 'log2']},\n",
       "             pre_dispatch='2*n_jobs', refit=True, return_train_score=False,\n",
       "             scoring=None, verbose=0)"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "grid_search2 = GridSearchCV(pipeline2, hyperparameters,cv=10)\n",
    "grid_search2.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.574288938609367\n",
      "3.7215504435277595\n"
     ]
    }
   ],
   "source": [
    "# 9. Evaluate model pipeline on test data\n",
    "pred3 = grid_search2.predict(X_test)\n",
    "print (r2_score(y_test, pred3))\n",
    "print (mean_squared_error(y_test, pred3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "RandomizedSearchCV(cv=10, error_score=nan,\n",
       "                   estimator=Pipeline(memory=None,\n",
       "                                      steps=[('columntransformer',\n",
       "                                              ColumnTransformer(n_jobs=None,\n",
       "                                                                remainder='drop',\n",
       "                                                                sparse_threshold=0.3,\n",
       "                                                                transformer_weights=None,\n",
       "                                                                transformers=[('num',\n",
       "                                                                               Pipeline(memory=None,\n",
       "                                                                                        steps=[('imputer',\n",
       "                                                                                                SimpleImputer(add_indicator=False,\n",
       "                                                                                                              copy=True,\n",
       "                                                                                                              fill_value=None,\n",
       "                                                                                                              missing_values=nan,\n",
       "                                                                                                              strategy='median',...\n",
       "                                                                    n_jobs=None,\n",
       "                                                                    oob_score=False,\n",
       "                                                                    random_state=None,\n",
       "                                                                    verbose=0,\n",
       "                                                                    warm_start=False))],\n",
       "                                      verbose=False),\n",
       "                   iid='deprecated', n_iter=10, n_jobs=4,\n",
       "                   param_distributions={'randomforestregressor__max_depth': [None,\n",
       "                                                                             5,\n",
       "                                                                             3,\n",
       "                                                                             1],\n",
       "                                        'randomforestregressor__max_features': ['auto',\n",
       "                                                                                'sqrt',\n",
       "                                                                                'log2']},\n",
       "                   pre_dispatch='2*n_jobs', random_state=0, refit=True,\n",
       "                   return_train_score=False, scoring=None, verbose=0)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.model_selection import RandomizedSearchCV\n",
    "\n",
    "rand_search = RandomizedSearchCV(pipeline2, hyperparameters, random_state=0,cv=10,n_jobs=4)\n",
    "rand_search.fit(X_train, y_train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'randomforestregressor__max_depth': None,\n",
       " 'randomforestregressor__max_features': 'sqrt'}"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rand_search.best_params_"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5747917393481584\n",
      "3.717154977959299\n"
     ]
    }
   ],
   "source": [
    "# 9. Evaluate model pipeline on test data\n",
    "pred4 = rand_search.predict(X_test)\n",
    "print (r2_score(y_test, pred4))\n",
    "print (mean_squared_error(y_test, pred4))"
   ]
  }
 ],
 "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}