如何移除flextable中含全量NA值的2022列？

方法一：创建tabulator时动态排除全NA年份 #

先判断数据里的2022列是否全部为NA，再决定要不要在tabulator中定义该列，避免无效列生成：

# 检查Yr2022是否全为NA
is_2022_all_na <- all(is.na(test$Yr2022))

# 先构建基础的tabulator参数
tab_args <- list(
  x = test, 
  rows = c("Sex", "Measure", "AgeGroup"),
  columns = "timer",
  `2019` = as_paragraph(as_chunk(Yr2019, formatter = as.integer)), 
  `2020` = as_paragraph(as_chunk(Yr2020, formatter = as.integer)),
  `2021` = as_paragraph(as_chunk(Yr2021, formatter = as.integer))
)

# 只有当2022列不全为NA时，才添加该列的定义
if(!is_2022_all_na){
  tab_args[["2022"]] <- as_paragraph(as_chunk(Yr2022, formatter = as.integer))
}

# 生成tabulator和flextable
ct <- do.call(tabulator, tab_args)
ft <- as_flextable(ct)
ft

方法二：生成flextable后直接删除指定列 #

如果已经生成了包含2022列的flextable，直接用delete_columns()方法移除目标列：

ct <- tabulator(
  x = test, rows = c("Sex", "Measure", "AgeGroup"),
  columns = "timer",
  `2019` = as_paragraph(as_chunk(Yr2019, formatter = as.integer)), 
  `2020` = as_paragraph(as_chunk(Yr2020, formatter = as.integer)),
  `2021` = as_paragraph(as_chunk(Yr2021, formatter = as.integer)),
  `2022` = as_paragraph(as_chunk(Yr2022, formatter = as.integer))
)
ft <- as_flextable(ct)
# 直接删除2022列
ft <- delete_columns(ft, j = "2022")
ft

要是需要批量删除所有全NA的列，可以先遍历列判断后再删除：

# 获取flextable的所有列名
col_names <- colnames(ft)

# 筛选出所有全为NA的列
na_cols <- sapply(col_names, function(col){
  all(is.na(ft$body$dataset[[col]]))
})

# 删除这些无效列
ft <- delete_columns(ft, j = names(na_cols[na_cols]))
ft

内容的提问来源于stack exchange，提问作者Fred