如何将含数值与字符的嵌套列拆分为单一字符列？

问题描述 #

使用arsenal包的comparedf函数对比两个数据框生成差异报告时，需要拆分结果中的嵌套列（values.x和values.y）以导出到Excel，但这两列同时包含数值（double）和字符（character）类型的嵌套元素，直接用tidyr::unnest会报错：

Can't combine `x[[1]]` <double> and `x[[4]]` <character>.

示例数据 #

df1 <- data.frame(id = c("a", "b", "c", "d","e"),
                  var = c(1, 2, 3, 4, 5),
                  var2 = c("A","B","C","A","B"))
df2 <- data.frame(id = c("a", "b", "c", "d","e"),
                  var =c(1,3,4,2,5),
                  var2 = c("C","A","C","A","B"))

对比代码及报错操作 #

library(arsenal)
summary(comparedf(df1, df2, by="id"))

# 尝试拆分嵌套列时报错
library(tidyr)
vec1 <- summary(comparedf(df1, df2, by="id"))
diftable <- vec1$diffs.table |> unnest(cols = c(values.x, values.y))

解决方案 #

核心思路是先将嵌套列中的所有元素统一转为字符类型，再执行拆分操作，避免类型冲突。可以借助purrr::map函数遍历嵌套列表的每个元素并转换类型：

library(arsenal)
library(tidyr)
library(purrr)

# 生成示例数据
df1 <- data.frame(id = c("a", "b", "c", "d","e"),
                  var = c(1, 2, 3, 4, 5),
                  var2 = c("A","B","C","A","B"))
df2 <- data.frame(id = c("a", "b", "c", "d","e"),
                  var =c(1,3,4,2,5),
                  var2 = c("C","A","C","A","B"))

# 获取对比结果
vec1 <- summary(comparedf(df1, df2, by="id"))
diftable <- vec1$diffs.table

# 将嵌套列的每个元素转为字符型
diftable <- diftable |>
  mutate(
    values.x = map(values.x, as.character),
    values.y = map(values.y, as.character)
  )

# 拆分嵌套列
diftable_unnested <- unnest(diftable, cols = c(values.x, values.y))

执行后，diftable_unnested中的values.x和values.y会变成统一的字符类型列，可直接导出到Excel。

内容的提问来源于stack exchange，提问作者Victor Shin