跨多列的未配对 t 检验的 R 代码

Question

我希望在以下代码中调整 p 值（fdr 方法）：

>lstOut <- combn(unique(df1$ATTRIBUTE_Group), 2, FUN = function(x) {+ dat1 <- subset(df1, ATTRIBUTE_Group == x[1])+ dat2 <- subset(df1, ATTRIBUTE_Group == x[2])+ Map(function(x, y) tryCatch(t.test(x, y,var.equal=TRUE), error = function(e) NA), dat1[-1], dat2[-1])+ }, simplify = FALSE)
>names(lstOut) <- combn(as.character(unique(df1$ATTRIBUTE_Group)), 2, FUN = paste, collapse="_")
>out <- map_depth(lstOut, .depth = 2, tidy)
>out2 <- map_dfr(out, ~bind_rows(.x, .id = 'colname'), .id = 'classCompare')

非常感谢， L.

Answer 1

我们可以使用p.adjust

library(dplyr)
out3 <- out2 %>%
              group_by(colname) %>%
              mutate(padj = p.adjust(p.value, method = "BH"))

跨多列的未配对 t 检验的 R 代码

R code for unpaired t-test across multiple columns

statistics

r

t-test