无法识别地图函数内的对象

问题描述

我有一个嵌套的数据框，我想获取子数据框中每一列不为 NA 的行数，如下(简化版):

I have a nested dataframe and I would like to get the numbers of rows which is not NA for each column in the child dataframe as follows (simplified version):

df <- list(
  tibble(a=c(1, 2, NA_real_, 4, 5), b=c(2, NA_real_, 5, NA_real_, 3)),
  tibble(a=c(NA_real_, 2, 3, NA_real_, 5), b=c(NA_real_, NA_real_, NA_real_, 1, 3))
) %>%
  tibble(x=1:2, y=.)

res <- df %>%
  mutate(z=map(y, function(dat){
    c(a="a", b="b") %>% map_int(function(col){
      dat %>% filter(!is.na(!!sym(col))) %>% nrow()
    }) %>% enframe()
  }))

当我运行代码时，我收到一条错误消息:错误:只有字符串可以转换为符号.函数内部的 col 好像无法识别.
为什么会出现这种行为?

When I run the code, I get an error that says: Error: Only strings can be converted to symbols. It seems that col inside the function is not recognized.
Why does this behavior occur?

推荐答案

如果你的数据叫做 df，你可以这样做:

If your data is called df, you could do :

library(dplyr)
library(purrr)

df1 <- df %>% mutate(res = map(y, ~.x %>% summarise_all(~sum(!is.na(.)))))

df1$res
#[[1]]
# A tibble: 1 x 2
#      a     b
#  <int> <int>
#1     4     3

#[[2]]
# A tibble: 1 x 2
#      a     b
#  <int> <int>
#1     3     2

或者使用基础 R :

sapply(df$y, function(x) colSums(!is.na(x)))

这篇关于无法识别地图函数内的对象的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持！