决策树的代码实现

#!usr/bin/env python
#-*- coding:utf-8 _*-
'''
@author:Administrator
@file: decision_tree.py
@time: 2020-05-11 下午 4:03
'''
import pandas as pd
from sklearn.feature_extraction import DictVectorizer
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier, export_graphviz


def decisionTree():
    '''
    决策树对数据预测
    :return:
    '''
######1.读取所有数据
    tree_data=pd.read_csv("../data/tree_data.csv");
    #tree_data=tree_data.head(5);
    print(tree_data);
    #获取指定列
    print(tree_data['pclass']);
#####2. 处理数据，找出特征值和目标值
     #获取多个特征时[[x1,x2,x3]]
    x = tree_data[['pclass',"age","sex"]]
    #print(x);
    y = tree_data['survived']
    #print(x);
    #print(y);
#####3. 缺失值处理,使用平均数进行插补
    print(x['age'].mean())
    ####这个写法无法完成插补，报错，还是nan
    #x['age'].fillna(x['age'].mean(), inplace=True)
    ######这个办法完成了插补
    x.fillna({'age': x['age'].mean()}, inplace=True);
    print(x['age']);

#####4. 分隔数据集
    x_train,x_test,y_train,y_test=train_test_split(x,y,test_size=0.25);

#####5.进行处理（特征工程）特征-》类别-》one_hot编码
    dict = DictVectorizer(sparse=False)
    #orient =’records’ ，转化后是 list形式：[{column(列名) : value(值)}……{column:value}];
    #详细介绍参考：https://www.lizenghai.com/archives/51607.html
    x_train = dict.fit_transform(x_train.to_dict(orient="records"))
    print(dict.get_feature_names())
    x_test = dict.transform(x_test.to_dict(orient="records"))

    print(x_train)
########6.用决策树进行预测
    dec = DecisionTreeClassifier()
    dec.fit(x_train, y_train)
######7.模型评估
    # # 预测准确率
    print("预测的准确率：", dec.score(x_test, y_test))
    print("预测：",dec.predict(x_test));
  # # 导出决策树的结构
    export_graphviz(dec, out_file="./tree.dot", feature_names=['年龄', 'pclass=1st', '女性', '男性'])
decisionTree();

https://blog.youkuaiyun.com/u011066470/article/details/106060002

该看地14集