编写自定义函数时将列名传递给 dplyr::coalesce()

Question

我正在尝试编写一个函数来包装 dplyr::coalesce()，并将接收数据对象和列名以合并。到目前为止，我的尝试都失败了。

示例数据

library(dplyr)

df <-
  data.frame(col_a = c("bob", NA, "bob", NA, "bob"), 
                 col_b = c(NA, "danny", NA, NA, NA), 
                 col_c = c("paul", NA, NA, "paul", NA))

##   col_a col_b col_c
## 1   bob  <NA>  paul
## 2  <NA> danny  <NA>
## 3   bob  <NA>  <NA>
## 4  <NA>  <NA>  paul
## 5   bob  <NA>  <NA>

在编写自定义函数时存根

coalesce_plus_1 <- function(data, vars) {

  data %>%
    mutate(coalesced_col = coalesce(!!! rlang::syms(tidyselect::vars_select(names(.), vars))))

}

coalesce_plus_2 <- function(data, vars) {
  
  data %>%
    mutate(coalesced_col = coalesce(!!! rlang::syms(vars)))
  
}

coalesce_plus_3 <- function(data, vars) {
  
  data %>%
    mutate(coalesced_col = coalesce({{ vars }}))
  
}

结果...

coalesce_plus_1()

df %>%
  coalesce_plus_1(data = ., vars = c(col_a, col_b, col_c))

Error: object 'col_a' not found.

但是：

df %>%
  coalesce_plus_1(data = ., vars = all_of(starts_with("col")))

##   col_a col_b col_c coalesced_col
## 1  <NA>  <NA>  paul          paul
## 2  <NA> danny  <NA>         danny
## 3   bob  <NA>  <NA>           bob
## 4  <NA>  <NA>  paul          paul
## 5   bob  <NA>  <NA>           bob

coalesce_plus_2()

df %>%
  coalesce_plus_2(data = ., vars = c(col_a, col_b, col_c))

Error in lapply(.x, .f, ...) : object 'col_a' not found

还有

df %>%
  coalesce_plus_2(data = ., vars = all_of(starts_with("col")))

Error: starts_with() must be used within a selecting function. i See https://tidyselect.r-lib.org/reference/faq-selection-context.html. Run rlang::last_error() to see where the error occurred.

coalesce_plus_3()

df %>%
  coalesce_plus_3(data = ., vars = c(col_a, col_b, col_c))

Error: Problem with mutate() input coalesced_col. x Input coalesced_col can't be recycled to size 5. i Input coalesced_col is coalesce(c(col_a, col_b, col_c)). i Input coalesced_col must be size 5 or 1, not 15.

还有

df %>%
  coalesce_plus_3(data = ., vars = all_of(starts_with("col")))

Error: Problem with mutate() input coalesced_col. x starts_with() must be used within a selecting function. i See https://tidyselect.r-lib.org/reference/faq-selection-context.html. i Input coalesced_col is coalesce(all_of(starts_with("col"))).

底线

我如何为 coalesce() 编写一个自定义函数，它将接受一个数据对象和特定的列名来合并，允许特定的命名，例如 c(col_a, col_b, col_c) 和辅助函数，例如 starts_with("col") 在函数的 vars 参数中？

Answer 1

这是一个简单的实现，它只会 return select 列，但可以相当容易地扩展以保留所有列（我会 bind_cols 最后将它们重新打开。 ..).

这很简单，因为我们依靠 select 为我们完成工作，正如 Implementing tidyselect vignette

开头所建议的那样

# edited to keep all columns
coalesce_df = function(data, ...) {
  data %>%
    select(...) %>%
    transmute(result = invoke(coalesce, .)) %>%
    bind_cols(data, .)
}



df %>%
   coalesce_df(everything())
#   col_a col_b col_c result
# 1   bob  <NA>  paul    bob
# 2  <NA> danny  <NA>  danny
# 3   bob  <NA>  <NA>    bob
# 4  <NA>  <NA>  paul   paul
# 5   bob  <NA>  <NA>    bob

df %>% coalesce_df(col_a, col_b)
#   col_a col_b col_c result
# 1   bob  <NA>  paul    bob
# 2  <NA> danny  <NA>  danny
# 3   bob  <NA>  <NA>    bob
# 4  <NA>  <NA>  paul   <NA>
# 5   bob  <NA>  <NA>    bob

Answer 2

实际上，您的第一个函数可以工作，只需将 vars 写为一个字符即可。看：

df %>% coalesce_plus_1(data = ., vars = c("col_a","col_b","col_c"))

这是另一个不错的选择：

library(dplyr)

df <- data.frame(col_a = c("bob", NA, "bob", NA, "bob"), 
                 col_b = c(NA, "danny", NA, NA, NA), 
                 col_c = c("paul", NA, NA, "paul", NA))

coalesce_plus <- function(data,vars){
      x <- as.list(select(data,vars))
      data.frame(data, coalesced_col=coalesce(!!!x))
}

df %>% coalesce_plus(data = ., vars = c("col_a","col_b","col_c"))

编写自定义函数时将列名传递给 dplyr::coalesce()

Pass column names to dplyr::coalesce() when writing a custom function

r

function

dplyr

tidyselect

示例数据

在编写自定义函数时存根

结果...

底线