Problem

Question

如无现有栏目,则以下职能在数据框架中的第一个现有栏目即。

library(dplyr, warn.conflicts = FALSE)
#> Warning: package  dplyr  was built under R version 4.2.3

# `dplyr::coalesce` return fist non-missing, but these columns must exist

return_exist_var <- function(.var1, .var2, .fill = NA, .data) {
  # get the . from magrittr::%>%
  .data <- get(".", parent.frame())
  stopifnot(is.data.frame(.data))
  .var1 <- rlang::enexpr(.var1)
  .var2 <- rlang::enexpr(.var2)
  if (rlang::as_string(.var1) %in% names(.data)) {
    return(rlang::parse_expr(rlang::as_string(.var1)))
  } else if (rlang::as_string(.var2) %in% names(.data)) {
    return(rlang::parse_expr(rlang::as_string(.var2)))
  } else {
    return(.fill)
  }
}

df <- data.frame(
  ID = c("ID01", "ID01", "ID02", "ID02"),
  v1 = c("a", "b", "c", NA),
  v2 = c("A", "B", "C", "C")
)

return first existing column

df %>% dplyr::mutate(NEW = !!return_exist_var(v1))
#>     ID   v1 v2  NEW
#> 1 ID01    a  A    a
#> 2 ID01    b  B    b
#> 3 ID02    c  C    c
#> 4 ID02 <NA>  C <NA>

return first existing column in two column

df %>% dplyr::mutate(NEW = !!return_exist_var(v2, v1))
#>     ID   v1 v2 NEW
#> 1 ID01    a  A   A
#> 2 ID01    b  B   B
#> 3 ID02    c  C   C
#> 4 ID02 <NA>  C   C

return second existing column in two column

df %>% dplyr::mutate(NEW = !!return_exist_var(v1_x, v2))
#>     ID   v1 v2 NEW
#> 1 ID01    a  A   A
#> 2 ID01    b  B   B
#> 3 ID02    c  C   C
#> 4 ID02 <NA>  C   C

return NA if two column not eixst

df %>% dplyr::mutate(NEW = !!return_exist_var(v1_X, v2_X))
#>     ID   v1 v2 NEW
#> 1 ID01    a  A  NA
#> 2 ID01    b  B  NA
#> 3 ID02    c  C  NA
#> 4 ID02 <NA>  C  NA

work well on grouped data frame

df %>%
  dplyr::group_by(ID) %>%
  dplyr::mutate(NEW = !!return_exist_var(v1))
#> # A tibble: 4 × 4
#> # Groups:   ID [2]
#>   ID    v1    v2    NEW  
#>   <chr> <chr> <chr> <chr>
#> 1 ID01  a     A     a    
#> 2 ID01  b     B     b    
#> 3 ID02  c     C     c    
#> 4 ID02  <NA>  C     <NA>

Problem

如何更新职能,然后将职能称为“无<条码>!

df %>% dplyr:mutate(NEW=回报_exist_var(v1)

Answer 1

下面是你职能的简化版本,其依据是使用<代码>coalesce(>>和any_of(>处理缺失的var。由于any_of()要求具有病媒特性,因此,在使用前必须加以区别。

library(dplyr)

return_exist_var <- function(...) {
  var_list <-  match.call(expand.dots = FALSE)$...
  do.call(coalesce, c(NA, pick(any_of(sapply(var_list, deparse)))))
}

df %>%
  dplyr::mutate(NEW = return_exist_var(v1))

    ID   v1 v2  NEW
1 ID01    a  A    a
2 ID01    b  B    b
3 ID02    c  C    c
4 ID02 <NA>  C <NA>

df %>%
  dplyr::mutate(NEW = return_exist_var(v2, v1))

    ID   v1 v2 NEW
1 ID01    a  A   A
2 ID01    b  B   B
3 ID02    c  C   C
4 ID02 <NA>  C   C

df %>%
  dplyr::mutate(NEW = return_exist_var(v1_x, v2))

    ID   v1 v2 NEW
1 ID01    a  A   A
2 ID01    b  B   B
3 ID02    c  C   C
4 ID02 <NA>  C   C

df %>% 
  dplyr::mutate(NEW = return_exist_var(v1_X, v2_X))

    ID   v1 v2 NEW
1 ID01    a  A  NA
2 ID01    b  B  NA
3 ID02    c  C  NA
4 ID02 <NA>  C  NA

# Works if data is grouped
df %>%
  mutate(NEW = return_exist_var(v1), .by = ID)

    ID   v1 v2  NEW
1 ID01    a  A    a
2 ID01    b  B    b
3 ID02    c  C    c
4 ID02 <NA>  C <NA>

return first existing column

return first existing column in two column

return second existing column in two column

return NA if two column not eixst

work well on grouped data frame

Problem

友情链接