Gtsummary error when using tbl_merge: "Can't convert from `y$modify_stat_N` <double> to `x$modify_stat_N` <integer> due to loss of precision."

I have a survey with missing responses and calculated weights to account for attrition. I want to be able to display both tables without weights and with weights side-by-side for comparison. I therefore want to merge the tables created with tbl_summary() and tbl_svysummary() using tbl_merge(), but end up with an error:

Error in dplyr::rows_update(): ! Can't convert from y$modify_stat_N \<double\> to x$modify_stat_N \<integer\> due to loss of precision. • Locations: 1, 2

Below is a reproducible example. I have already found that the error does not occur when using the by argument in tbl_summary() or tbl_svysummary().

Many thanks

set.seed(123)
library(tidyverse)
library(gtsummary)
library(srvyr)

num_rows <- 10000

item1 <- sample(c("Never", "Sometimes", "All the time"), num_rows, replace = TRUE)
item2 <- sample(c("Never", "Sometimes", "All the time"), num_rows, replace = TRUE)
item3 <- sample(c("Never", "Sometimes", "All the time"), num_rows, replace = TRUE)

surveystatus <- sample(c("Respondent", "Non-respondent"), num_rows, replace = TRUE)

gender <- sample(c("Male", "Female"), num_rows, replace = TRUE)

weight <- rnorm(num_rows, mean = 0, sd = 1)

toy_data <- data.frame(item1, item2, item3, gender, weight, surveystatus)

  # Create a survey object
toy_dataw <- toy_data %>%
  srvyr::as_survey_design(weights = weight)

####
  # Tbl merge that works ! :D
t1 <- toy_data %>%
  filter(surveystatus == "Respondent") %>%
  select(item1, item2, item3, gender) %>%
  tbl_summary(by = gender)

t2 <- toy_dataw %>%
  srvyr::filter(surveystatus == "Respondent") %>%
  srvyr::select(item1, item2, item3, gender) %>%
  tbl_svysummary(by = gender)

tbl_merge(tbls = list(t1, t2))

  # Tbl merge that does not work ! :(
t3 <- toy_data %>%
  filter(surveystatus == "Respondent") %>%
  select(item1, item2, item3) %>%
  tbl_summary()

t4 <- toy_dataw %>%
  srvyr::filter(surveystatus == "Respondent") %>%
  srvyr::select(item1, item2, item3) %>%
  tbl_svysummary()

tbl_merge(tbls = list(t3, t4))

it looks like you've identified a bug. I'll address it in the next release. In the meantime, here's a workaround. Happy Programming

library(survey) |> suppressPackageStartupMessages()
library(gtsummary)
#> #BlackLivesMatter
data(api)

#stratified sample
dstrat <- svydesign(id=~1,strata=~stype, weights=~pw, data=apistrat, fpc=~fpc)

t1 <- tbl_summary(apistrat, include = api00)
t2 <- tbl_svysummary(dstrat, include = api00)

# convert `modify_stat_N` and `modify_stat_n` to a double
for (int_col in c("modify_stat_N", "modify_stat_n")) {
  t1$table_styling$header[[int_col]] <-
    t1$table_styling$header[[int_col]] |> as.numeric()
}

final_tbl <-
  list(t1, t2) |> 
  tbl_merge()

^{Created on 2024-04-14 with reprex v2.1.0}