Introduction to Emsembling/Stacking in Python

泰坦尼克号

https://www.kaggle.com/arthurtok/introduction-to-ensembling-stacking-in-python

代码理解

1. Preprocessing

1.库加载

import plotly.offline as py
py.init_notebook_mode(connected=True)
import plotly.graph_objs as go
import plotly.tools as tls

import warnings
warnings.filterwarnings('ignore')
import xgboost as xgb

Going to use these 5 base models for the stacking

from sklearn.ensemble import (RandomForestClassifier, AdaBoostClassifier,
GradientBoostingClassifier, ExtraTreesClassifier)
from sklearn.svm import SVC
from sklearn.cross_validation import KFold

1. 对于feature中含有缺失值（缺失值比例较大时），把缺失的和未缺失的来看

    train['Has_Cabin'] = train["Cabin"].apply(lambda x: 0 if type(x) == float else 1)
   

2. 分桶处理

    train['CategoricalFare'] = pd.qcut(train['Fare'], 4)
    dataset.loc[ dataset['Fare'] <= 7.91, 'Fare'] 						        = 0
    dataset.loc[(dataset['Fare'] > 7.91) & (dataset['Fare'] <= 14.454), 'Fare'] = 1
    dataset.loc[(dataset['Fare'] > 14.454) & (dataset['Fare'] <= 31), 'Fare']   = 2
    dataset.loc[ dataset['Fare'] > 31, 'Fare'] 							        = 3
    dataset['Fare'] = dataset['Fare'].astype(int)
   

3. 填充缺失值（通过均值和标准差来填充）

    for dataset in full_data:
        age_avg = dataset['Age'].mean()
        age_std = dataset['Age'].std()
        age_null_count = dataset['Age'].isnull().sum()
        age_null_random_list = np.random.randint(age_avg - age_std, age_avg + age_std, size=age_null_count)
        dataset['Age'][np.isnan(dataset['Age'])] = age_null_random_list
        dataset['Age'] = dataset['Age'].astype(int)
   

4. 替换

   dataset['Title'] = dataset['Title'].replace('Mlle', 'Miss')
   

5. 编码mapping

   dataset['Sex'] = dataset['Sex'].map( {'female': 0, 'male': 1} ).astype(int)
   

6. 热力图（去除冗余变量）

    colormap = plt.cm.RdBu
    plt.figure(figsize=(14,12))
    plt.title('Pearson Correlation of Features', y=1.05, size=15)
    sns.heatmap(train.astype(float).corr(),linewidths=0.1,vmax=1.0, 
                square=True, cmap=colormap, linecolor='white', annot=True)
   

2. Ensembling & Stacking models

# Some useful parameters which will come in handy later on
ntrain = train.shape[0]
ntest = test.shape[0]
SEED = 0 # for reproducibility
NFOLDS = 5 # set folds for out-of-fold prediction
kf = KFold(ntrain, n_folds= NFOLDS, random_state=SEED)

# Class to extend the Sklearn classifier
class SklearnHelper(object):
    def __init__(self, clf, seed=0, params=None):
        params['random_state'] = seed
        self.clf = clf(**params)

    def train(self, x_train, y_train):
        self.clf.fit(x_train, y_train)

    def predict(self, x):
        return self.clf.predict(x)
    
    def fit(self,x,y):
        return self.clf.fit(x,y)
    
    def feature_importa