分享

训练集、测试集(train_test_split)

 印度阿三17 2021-02-02

如果拿所有原始数据来训练,存在的问题:

  1. 模型很差无法调整;

  2. 真实环境难以拿到真实 label;

所以将数据区分为 训练数据 和 测试数据(train test split);
将训练数据来训练模型;然后用测试数据测试模型;

使用这种方式也存在问题;


python 原生分离 iris 数据集

import numpy as np
from sklearn import datasets
import matplotlib.pyplot as plt
 
iris = datasets.load_iris()
 
X = iris.data 
y = iris.target
 
X
''' 
    (array([[5.1, 3.5, 1.4, 0.2],
            [4.9, 3. , 1.4, 0.2],  ...
            [6.2, 3.4, 5.4, 2.3],
            [5.9, 3. , 5.1, 1.8]])
'''

y 
'''
    array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
           0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
           0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
           1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
           1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
           2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
           2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2])

'''

X.shape, y.shape #((150, 4), (150,))
 
# shuffle 
shuffle_indexes = np.random.permutation(len(X))  # 0--len(X) 的随机排列
shuffle_indexes
# array([ 22,   4, 142,  24,   7, 146,  ... 9,  95, 130,  29, 124])
 
test_ratio = 0.2
test_size = int(len(X) * test_ratio)
test_size # 30
 
test_indexes = shuffle_indexes[:test_size]
train_indexes = shuffle_indexes[test_size:]
 
test_indexes  
''' 
    array([ 22,   4, 142,  24,   7, 146,  70,  77, 144,  14,  40, 119,  46, 85,  74,  87,  86,  60,  91, 120,  78,  45,  65, 105, 113,  39, 83,  80, 134,  16])
'''

X_train = X[train_indexes]
y_train = y[train_indexes]

X_test = X[test_indexes]
y_test = y[test_indexes]
  
X_test.shape, X_train.shape # ((30, 4), (120, 4))

封装 train_test_split 函数

def train_test_split(X, y, test_ratio=0.2, seed=None):
    
    assert x.shape[0] == y.shape[0], "the size of X must be equal to the size of y"
    assert 0.0 <= test_ratio <= 1.0, "test_ ration must be valid"
    if seed:
        np.random.seed(seed)
        
    shuffle_indexes = np.random.permutation(len(X))  
    test_size = int(len(X) * test_ratio)
    
    test_indexes = shuffle_indexes[:test_size]
    train_indexes = shuffle_indexes[test_size:]

    X_train = X[train_indexes]
    y_train = y[train_indexes]

    X_test = X[test_indexes]
    y_test = y[test_indexes]
    
    return X_train, y_train, X_test, y_test

sklearn 中的 train_test_split

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y)

train_test_split(*arrays, **options)

X_train, X_test, y_train, y_test = train_test_split( X, y, test_size=0.33, random_state=42)


来源:https://www./content-4-842901.html

    本站是提供个人知识管理的网络存储空间,所有内容均由用户发布,不代表本站观点。请注意甄别内容中的联系方式、诱导购买等信息,谨防诈骗。如发现有害或侵权内容,请点击一键举报。
    转藏 分享 献花(0

    0条评论

    发表

    请遵守用户 评论公约

    类似文章 更多