数据挖掘学习（二）

1、train.csv
这里采用的数据集来自于Kaggle | Allstate Claims Severity比赛，这里的训练集如下所示，有116个离散特征（cat1-cat116）,14个连续特征（cont1 -cont14）

2、xgboost代码分析

import numpy as np  
import pandas as pd  
import xgboost as xgb  
import operator  
import matplotlib.pyplot as plt  

# 创建 features 矩阵，用于绘制 features importance (对应的分数)
# 生成 i feat q 格式的数据（i++）

def ceate_feature_map(features):  
    outfile = open('xgb.fmap', 'w')  
    i = 0  
    for feat in features:  
        outfile.write('{0}\t{1}\tq\n'.format(i, feat))  
        i = i + 1  
    outfile.close()  


if __name__ == '__main__':  
    train = pd.read_csv("../input/train.csv")  

# 离散特征用字符串表示，先要对其进行数值化
# 其中train["cat1"] 为 index+values
# pd.factorize(train[column].values , sort=True):(array([0, 0, 0, ..., 0, 0, 1]), array(['A', 'B'], dtype=object))

    cat_sel = [n for n in train.columns if n.startswith('cat')]  #类别特征数值化  
    for column in cat_sel:  
        train[column] = pd.factorize(train[column].values , sort=True)[0] + 1  

    params = {  
        'min_child_weight': 100,  
        'eta': 0.02,  
        'colsample_bytree': 0.7,  
        'max_depth': 12,  
        'subsample': 0.7,  
        'alpha': 1,  
        'gamma': 1,  
        'silent': 1,  
        'verbose_eval': True,  
        'seed': 12  
    }  
    rounds = 10  
    # label y
    y = train['loss']  
    # 去除不相关向量
    X = train.drop(['loss', 'id'], 1)  

    xgtrain = xgb.DMatrix(X, label=y)  
    bst = xgb.train(params, xgtrain, num_boost_round=rounds)  

    # 类别特征数值化
    # 过滤非需要的的target
    features = [x for x in train.columns if x not in ['id','loss']]  
    ceate_feature_map(features)  

    importance = bst.get_fscore(fmap='xgb.fmap') 
    # 从dist 转为 list 
    importance = sorted(importance.items(), key=operator.itemgetter(1))  
  # 给一个list 转化为带有 columns的DataFrame
    df = pd.DataFrame(importance, columns=['feature', 'fscore'])  
    # 比值化
    df['fscore'] = df['fscore'] / df['fscore'].sum()  
    df.to_csv("./feat_importance.csv", index=False)  

    plt.figure()  
    df.plot(kind='barh', x='feature', y='fscore', legend=False, figsize=(6, 10))  
    plt.title('XGBoost Feature Importance')  
    plt.xlabel('relative importance')  
    plt.show()