我想根据缺失数据百分比是否超过10% 来在列名上添加星号。举例来说，在这个包含3 个变量的表中，我希望特征列显示为： "Age", "Result*" 和 "Group*"。

df <- data.frame(
  Age = c(1:90, rep(NA, 10)),                # 10% Missing
  Result = c(1:80, rep(NA, 20)),             # 20% Missing
  Group = sample(c("A", "B", NA), 100, replace=TRUE) # Missingness varies
)

我正在使用R 语言中的一个很棒的gtsummary包，以及tbl_summary函数。到目前为止，我已经能够使用add_n计算缺失数据的百分比，但似乎无法让下一步生效。如有任何建议，敬请指教！

library(gtsummary)

table <- tbl_summary(df) |> 
  add_n("{p_miss}") 

table |> 
  modify_table_styling(columns = label, rows = row_type == "label", label = ifelse(n > 10, paste0(label, "*"), label))

解决方案

Here's an example adding an asterix and also adding a note via a proper footnote as well.

library(gtsummary)

table <- trial |> 
  tbl_summary(
    by = trt,
    include = c(marker, age),
    missing = "no"
  ) |> 
  add_n("{p_miss}", col_label = "**% Missing**")

# create a vector of variable names with missing rate >5% (there are plenty of ways to get this and this is just one example)
variables_big_missing <-
  gather_ard(table)$add_n |> 
  cards::unlist_ard_columns() |> 
  dplyr::filter(stat_name == "p_miss", stat > 0.05) |> 
  dplyr::pull(variable)


table_final <- table |> 
  # add a star for variables with missing rate larger than %5
  modify_table_body(
    ~ .x |> 
      dplyr::mutate(label = ifelse(variable %in% .env$variables_big_missing & row_type == "label", paste0(label, "*"), label))
  ) |> 
  # or you can do it via a proper footnote
  modify_footnote_body(
    footnote = "Missing rate > 5%",
    columns = "label",
    rows = variable %in% .env$variables_big_missing & row_type == "label"
  )

创建于2026-06-07，使用 reprex v2.1.1