泰坦尼克乘客存活状况(决策树案例)

import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction import  DictVectorizer
from sklearn.tree import DecisionTreeClassifier
from sklearn.tree import export_graphviz
from sklearn.ensemble import RandomForestClassifier


def descsion():
    # 获取数据, 提取特征值和目标值
    Titanic_data = pd.read_csv("http://biostat.mc.vanderbilt.edu/wiki/pub/Main/DataSets/titanic.txt")

    # 打印字段名
    print(Titanic_data.columns)

    # 分割出特定的字段(社会阶层, 年龄, 性别)对生存率的影响
    titanic_x = Titanic_data[["pclass", "age", "sex"]]
    titanic_y = Titanic_data[["survived"]]
    # 处理缺失值
    titanic_x["age"].fillna(titanic_x['age'].mean(), inplace=True)

    # 进行数据的分割
    x_train, x_test, y_train, y_test = train_test_split(titanic_x, titanic_y, train_size=0.25)

    # 对特征们进行字典特征抽取
    dict = DictVectorizer(sparse=False)

    x_train = dict.fit_transform(x_train.to_dict(orient="records"))
    x_test = dict.transform(x_test.to_dict(orient="records"))

    # 查看抽取后特征的名字
    feature_names = dict.get_feature_names()
    print(feature_names)


    # 进行决策树预测(可选:限制决策树最大深度为10)
    my_decision_tree = DecisionTreeClassifier(max_depth=10)
    my_decision_tree.fit(x_train, y_train)

    print("单棵决策树预测的准确率为:", my_decision_tree.score(x_test, y_test))

    # 将树的结构保存到本地
    export_graphviz(my_decision_tree, "./my_decision_tree.dot", feature_names = feature_names)

    """
    将dot文件装换为png的方法
    在本机安装graphviz ubuntu版安装: sudo apt install graphviz   mac版安装: brew install graphviz
    然后运行命令: dot -Tpng my_decision_tree.dot -o my_decision_tree.png
    生成png格式图片my_desion_tree.png
    """
    # 随机树森林算法, 建立20棵数, 树的最大深度为15
    rf = RandomForestClassifier(n_estimators=21, max_depth=20)
    rf.fit(x_train, y_train)
    print("随机数森林预测的准确率为:", rf.score(x_test, y_test))

if __name__ == '__main__':
    descsion()

sudo apt install graphviz

brew install graphviz

dot -Tpng my_decision_tree.dot -o my_decision_tree.png