r - 从事后测试中获得整洁的输出

Question

考虑这个数据框dat1:

dat1 <- data.frame(Region = rep(c("r1","r2"), each = 100),
                   State = rep(c("NY","MA","FL","GA"), each = 10),
                   Loc = rep(c("a","b","c","d","e","f","g","h"),each = 5),
                   ID = rep(c(1:10), each=2),
                   var1 = rnorm(200),
                   var2 = rnorm(200),
                   var3 = rnorm(200),
                   var4 = rnorm(200),
                   var5 = rnorm(200))

我有与上面创建的 dat1 类似的数据框。 Region、State 和 Loc 是每个观测值 ID 的分组变量，每个观测值进行 5 次测量var1:var5。对于每个分组变量，我对每个 var 进行单变量方差分析。当发现显着差异时，我使用 multcompView 包中的 TukeyHSD() 函数和 multcompLetters() 函数在组。由于我想对每个分组变量执行此操作，因此我尝试编写一个函数来防止自己重复和拼写错误。下面显示了我的情况:

library(tidyverse)
library(multcomp)
library(multcompView)

Tuk <- function(dat,groupvar,var){
  TUK <- TukeyHSD(aov(lm(get(var) ~ get(groupvar), data=dat)))
  names(TUK)[[1]] <- paste0(groupvar)
  lets<-multcompLetters(extract_p(TUK$groupvar))
  lets
}
#assuming all 5 vars were significant in the anovas, I would then run this for each grouping variable as follows:
vars <- paste0(names(dat1[,5:9]))
#by Region
lapply(vars, FUN=Tuk, dat=dat1, groupvar="Region")
#by State
lapply(vars, FUN=Tuk, dat=dat1, groupvar="State")
#by Loc
lapply(vars, FUN=Tuk, dat=dat1, groupvar="Loc")

代码在函数之外运行。该函数将创建模型，但我不知道如何格式化它，以便它识别 groupvar 对于 multcompLetters(extract_p()) 部分是什么？我该如何解决这个问题，以及如何让它输出一个整洁的表格的功能，该表格显示每个组以及我立即给出的每个变量的字母。例如，对于使用所有 5 个变量的 State 来说，它看起来像这样

     NY   MA   FL   GA
var1  a    ab   c    a
var2  a    ab   b    c
var3  a    c    ab   bc  
var4  ab   c    ab   ab 
var5  a    b    c     b

此外，是否有一种合理的方法可以使该函数生成显示 CLD 字母的组的箱线图(针对每个变量)？

Answer 1

假设情节确实是您正在寻找的内容，这是否会让您非常接近 var1 ~ State 的歌唱情节，Indrajeet 在构建这个包方面做得很好，我讨厌重新发明轮子。

dat1 <- data.frame(Region = rep(c("r1","r2"), each = 100),
                   State = rep(c("NY","MA","FL","GA"), each = 10),
                   Loc = rep(c("a","b","c","d","e","f","g","h"),each = 5),
                   ID = rep(c(1:10), each=2),
                   var1 = rnorm(200),
                   var2 = rnorm(200),
                   var3 = rnorm(200),
                   var4 = rnorm(200),
                   var5 = rnorm(200))

library(ggstatsplot)

ggbetweenstats(dat1, State, var1, 
               plot.type = "box", 
               pairwise.comparisons = TRUE, 
               pairwise.display = "everything")

#> Note: Shapiro-Wilk Normality Test for var1: p-value = 0.183
#> 
#> Note: Bartlett's test for homogeneity of variances for factor State: p-value = 0.373
#>