python随机抽取样本_python 机器学习之随机抽样-分层抽样

在机器学习中，我们通常将原始数据按照比例分割为“测试集”和“训练集”。python有现成的模块，完整代码如下：

import pandas as pd

import numpy as np

import matplotlib.pyplot as plt

import seaborn as sns

from sklearn.model_selection import train_test_split

# 随机抽样

def read_data(data_file,label,start,test_size,random_state):

#label  分类变量所在列号，如第二列是2；start，分类用的指标开始列号；random_state,随机种子

dataset = pd.read_csv(data_file)

train_data= dataset.iloc[:,(start-1):].values

train_target = dataset.iloc[:,(label-1):label]

train_x,test_x,train_y,test_y = train_test_split(train_data,train_target,test_size=test_size,random_state=random_state)

return train_x, train_y, test_x, test_y,dataset

if __name__ == '__main__':

data_file = "/data/for_python.csv"

out_dir="/data/models_pipli