问题中的第二个输出表显示，您正在使用.作为big.mark/千分频器，即最后一列“N = 9.150”。readr::parse_numbers()默认情况下期望,作为千分频器，并将.解析为十进制标记。

library(tidyverse)

parse_number("9.150")
#> [1] 9.15

字符串
我们可以通过在locale参数中将decimal_mark定义为,来改变这一点。

readr::parse_number("9.150", locale = readr::locale(decimal_mark = ","))
#> [1] 9150

型
我改变了你的例子数据，所以我们的总数超过1000，而里约热内卢将保持在1000以下。

data <- data.frame(
  anoobito = sample(2019:2023),
  NOME_MUNICIPIO = sample(
    c("Rio de Janeiro", "São Paulo", "Araruama", "Queimados"),
    230000,
    prob = c(.01, 1, 1, 1),
    replace = TRUE
  )
)

library(gtsummary)
theme_gtsummary_language("en", big.mark = ".", decimal.mark = ",")

型
如果我们使用您的原始代码，里约热内卢排在第一位，因为它的总价值被解析为高于其他地方的价值。

data %>%
  select(anoobito, NOME_MUNICIPIO) %>%
  tbl_summary(by = anoobito,
              label = c(NOME_MUNICIPIO ~ "")) %>%
  add_overall(last = T) %>%
  modify_table_body(fun = ~ arrange(., desc(parse_number(stat_0))))

型
x1c 0d1x的数据
当我们在parse_number()调用中包含locale = locale(decimal_mark = ",")时，我们得到了所需的行顺序。

data %>%
  select(anoobito, NOME_MUNICIPIO) %>%
  tbl_summary(by = anoobito,
              label = c(NOME_MUNICIPIO ~ "")) %>%
  add_overall(last = T) %>%
  modify_table_body(fun = ~ arrange(., desc(parse_number(
    stat_0, locale = locale(decimal_mark = ",")
  ))))

型

的

Till的解决方案很棒。这是另一个不使用modify_table_body()的版本，我有时觉得它更简单。编程愉快！

library(gtsummary)
#> #BlackLivesMatter
packageVersion("gtsummary")
#> [1] '1.7.2'
# set language to "pt"
theme_gtsummary_language("pt", big.mark = ".", decimal.mark = ",")
#> Setting theme `language: pt`

# simulate data
data <- 
  data.frame(
    anoobito = sample(2019:2023),
    NOME_MUNICIPIO = sample(c("Rio de Janeiro", "São Paulo", "Araruama", "Queimados"), 1000, replace = TRUE)
  )

tbl <-
  data |> 
  # build primary summary table
  tbl_summary(
    by = anoobito, 
    sort = list(NOME_MUNICIPIO = "frequency")
  ) |> 
  # add spanning header to the years
  modify_spanning_header(all_stat_cols() ~ "**Ano do Óbito**") |> 
  # add an overall column
  add_overall() |> 
  # udpate the headers, and put the Ns on a new line
  modify_header(label = '**Município de Residência**',
                all_stat_cols() ~ "**{level}**  \n N = {style_number(n)}") |> 
  # we don't want the header row, so remove it and undo the indentation
  remove_row_type(type = "header") |> 
  modify_table_styling(
    columns = label,
    text_format = "indent",
    undo_text_format = TRUE
  )

字符串
x1c 0d1x的数据
创建于2023-10-27带有reprex v2.0.2