Question

我们在R e中使用var.test()函数，例如：

T1<-rnorm(255,mean=1.432,sd=0.255)
T2<-rnorm(256,mean=1.485,sd=0.251)
var.test(T1,T2)

# F test to compare two variances
# 
# data:  T1 and T2
# F = 1.1027, num df = 254, denom df = 255, p-value = 0.436
# alternative hypothesis: true ratio of variances is not equal to 1
# 95 percent confidence interval:
#   0.8620164 1.4106568
# sample estimates:
#   ratio of variances 
# 1.102695

但是，当我们使用相同的数据重新运行测试时，会得到非常不同的结果，例如：

T1<-rnorm(255,mean=1.432,sd=0.255)
T2<-rnorm(256,mean=1.485,sd=0.251)
var.test(T1,T2)

# F test to compare two variances
# 
# data:  T1 and T2
# F = 0.79853, num df = 254, denom df = 255, p-value = 0.07334
# alternative hypothesis: true ratio of variances is not equal to 1
# 95 percent confidence interval:
#   0.6242396 1.0215441
# sample estimates:
#   ratio of variances 
# 0.7985297

为什么会这样？我们在做错什么吗？

我们有多个数据集需要分析，我们需要了解正在发生的事情。

Answer 1

要使分析可重复，可以使用set.seed，它指定R随机数生成器的种子。

set.seed(42)  # set seed
T1 <- rnorm(255, mean=1.432, sd=0.255)
T2 <- rnorm(256, mean=1.485, sd=0.251)
var.test(T1, T2)

# same seed - same result
set.seed(42)
T1 <- rnorm(255, mean=1.432, sd=0.255)
T2 <- rnorm(256, mean=1.485, sd=0.251)
var.test(T1, T2)

# different seed - different result 
set.seed(123)
T1 <- rnorm(255, mean=1.432, sd=0.255)
T2 <- rnorm(256, mean=1.485, sd=0.251)
var.test(T1, T2)

var.test的问题

1 个答案: