Библиотека Sklearn

import argparse, pandas as pd
from sklearn import preprocessing
from sklearn.pipeline import Pipeline
from sklearn.ensemble import RandomForestClassifier, GradientBoostingClassifier
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import f_regression
from sklearn.grid_search import GridSearchCV
 
def rf(X,y, xtest, ytest):
    anova_pre = SelectKBest(f_regression, k=8)
    model = Pipeline([
        ('inp', preprocessing.Imputer(strategy='mean', missing_values=-1)),
        ('anova', anova_pre),
        ('clf', GradientBoostingClassifier()),
        ])
    grid = GridSearchCV(model, {
        'inp__strategy': ['mean', 'median'],
        'clf__learning_rate': [0.5, 0.8,1],
        'clf__max_depth':[5, 7, None],
        'clf__n_estimators': [50, 100],
        }, cv=5, verbose=3)
    model = grid.fit(X, y)
    print("This is test: ", model.score(xtest, ytest))
 
def model1(title):
    df = pd.read_csv('./data/train.csv')
    df = df.drop(['Name', 'Ticket', 'Cabin'], axis=1)
    df['Gender']= df['Sex'].map({'female':0, 'male': 1}).astype(int)    
    df['Embarked'] = df['Embarked'].fillna('S')
    df["Age"] = df["Age"].fillna(df["Age"].median())
    df = pd.concat([df, pd.get_dummies(df['Embarked'], prefix='Embarked')], axis=1)
    df = df.drop(['Sex', 'Embarked'], axis=1)
    cols = df.columns.tolist()
    cols = [cols[1]] + cols[0:1] + cols[2:]
    df = df[cols]
    train_data = df.values
    model = RandomForestClassifier(n_estimators=100)
    model = model.fit(train_data[0:, 2:], train_data[0:,0]) 
    rf(train_data[0:, 2:], train_data[0:,0], train_data[0:, 2:], train_data[0:,0])
    """+++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++"""
    df_test = pd.read_csv('./data/test.csv')
    df_test = df_test.drop(['Name', 'Ticket', 'Cabin'], axis=1)
    df_test['Gender']= df_test['Sex'].map({'female':0, 'male': 1}).astype(int)
    df["Age"] = df["Age"].fillna(df["Age"].median())
    fare_means = df.pivot_table('Fare', index='Pclass', aggfunc='mean')
    df_test['Fare'] = df_test[['Fare', 'Pclass']].apply(lambda x: fare_means[x['Pclass']]
        if pd.isnull(x['Fare']) else x['Fare'], axis=1)
    df_test = pd.concat([df_test, pd.get_dummies(df_test['Embarked'], prefix='Embarked')], axis=1)
    df_test = df_test.drop(['Sex', 'Embarked'], axis=1)
    test_data = df_test.values
                       
def run():
    parser = argparse.ArgumentParser(description='Parsing model arguments')
    parser.add_argument('--rf', help='Random forest model', default=False, action='store_true')
    result = parser.parse_args()
    model1(result)
               
run()