生成漂亮的 table 后，有没有办法在 gtsummary 中重新调整变量？

Question

我正在尝试在创建 table 后重新调整 gtsummary 对象中特定变量的级别。默认情况下，参考级别位于顶部（正如预期的那样），但有时您会想要特定的顺序。为此 table 我希望 agecat“61-83”被推下。我用 modify_table 试过了，但我明白了 relevel 错误（agecat，ref =“0-20”）：未找到对象 'agecat'

library(gtsummary)
library(dplyr)

data(trial)

trial <- trial %>%
  mutate(agecat = case_when(between(age, 0, 20) ~ "0-20",
                            between(age, 21, 40) ~ "21-40",
                            between(age, 41, 60) ~ "41-60",
                            between(age, 61, 83) ~ "61-83"),
         agecat = as.factor(agecat),
         agecat = relevel(agecat, ref = "61-83"))

glm(response ~ trt + grade + agecat,
    data = trial, family = binomial) %>%
  tbl_regression(exponentiate = TRUE)

Answer 1

有两种方法可以解决这个问题：

gtsummary table 的打印本质上是一个数据框，带有使其美观的格式选项。您可以使用 modify_table_body() 和 arrange() 函数对基础数据框进行排序。

library(gtsummary)
library(tidyverse)
packageVersion("gtsummary")


trial2 <-
  trial %>%
  mutate(
    agecat = case_when(
      between(age, 0, 20) ~ "0-20",
      between(age, 21, 40) ~ "21-40",
      between(age, 41, 60) ~ "41-60",
      between(age, 61, 83) ~ "61-83"
    ) %>%
      factor(),
  )

glm(response ~ trt + grade + agecat,
    data = trial2 %>% mutate(agecat = relevel(agecat, ref = "61-83")), 
    family = binomial) %>%
  tbl_regression(exponentiate = TRUE) %>%
  modify_table_body(
    ~ .x %>%
      arrange(variable, desc(row_type == "label"), label == "61-83")
  )

第二种方法是使用contrasts() 指定应引用哪个组。 tbl_regression() 保留顺序。

contrasts(trial2$agecat) <-
  contr.treatment(levels(trial2$agecat),
                  base = which(levels(trial2$agecat) == "61-83")
  )

glm(response ~ trt + grade + agecat,
    data = trial2, 
    family = binomial) %>%
  tbl_regression(exponentiate = TRUE)

生成漂亮的 table 后，有没有办法在 gtsummary 中重新调整变量？

Is there a way to relevel a variable in gtsummary after generating the beautiful table?

r

gtsummary