R:如果加载了插入符号包，则 set.seed() 结果不匹配 [英] R: set.seed() results don't match if caret package loaded

问题描述

我在 R(版本:3.3.0)中使用 createFolds() 来创建训练/测试分区.为了使结果可重现，我使用了种子值为 10 的 set.seed().正如预期的那样，结果(生成的折叠)是可重现的.

I am using createFolds() in R (version: 3.3.0) to create train/test partitions. To make results reproducible, I used set.seed() with a seed value of 10. As expected, the results (generated folds) were reproducible.

但是一旦我在设置种子后加载了插入符号包.然后使用createFolds函数，发现创建的折叠不一样了(虽然还是可以重现的).

But once I loaded caret package just after setting the seed. And then used the createFolds function, I found that the created folds were different (although still reproducible).

具体来说，创建的折叠在以下两种情况下有所不同:

Specifically, the created folds differ in the following two cases:

情况 1:

library(caret)
set.seed(10)
folds=createFolds(y,k=5,returnTrain=TRUE)

情况 2:

set.seed(10)
library(caret)
folds=createFolds(y,k=5,returnTrain=TRUE)

其中 y 是一个向量.

为什么会发生这种情况?

Why could this be happening?

推荐答案

罪魁祸首是 ggplot2，它在您加载插入符号时附加.它定义了一个 .onAttach 函数:https://github.com/hadley/ggplot2/blob/master/R/zzz.r

The culprit is ggplot2, which is attached when you load caret. It defines an .onAttach function: https://github.com/hadley/ggplot2/blob/master/R/zzz.r

这个函数在包被附加时被调用，参见help("ns-hooks").并在其中调用 runif 从而推进 RNG 的状态.

This function is called when the package is attached, see help("ns-hooks"). And within it runif is called thereby advancing the state of the RNG.

这篇关于R:如果加载了插入符号包，则 set.seed() 结果不匹配的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！