如何遍歷 R 中的列？

Question

我正在嘗試清理我的代碼以清理丟失的數據。 我有一個包含 6 列的數據集，如果我像這樣單獨執行它們，則代碼可以工作：

mammographic_masses <- mammographic_masses %>%
  mutate(birad = replace(birad, birad== "na", NA)) %>%
  mutate(birad = replace(birad, birad== "N/A", NA))

但是當我嘗試在這樣的 for 循環中執行此操作時：

for (i in ncol(mammographic_masses)){
  print(class(mammographic_masses[[i]]))
  mammographic_masses <- mammographic_masses %>%
    mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "na", NA)) %>%
    mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "N/A", NA))
}

我收到一個錯誤：

Error: unexpected '=' in:
"  mammographic_masses <- mammographic_masses %>%
    mutate(mammographic_masses[[,i]] ="
>     mutate(mammographic_masses[[,i]] = replace(mammographic_masses[[,i]], mammographic_masses[[,i]] == "N/A", NA))
Error: unexpected '=' in "    mutate(mammographic_masses[[,i]] ="
> }
Error: unexpected '}' in "}"

我也在閱讀其他方式，例如應用等，但我想不出一種方法來循環每列

Answer 1

使用mutate_all代替循環。

library(dplyr)

mammographic_masses %>%
  mutate_all(function(x) {is.na(x) <- x %in% c("na", "N/A"); x})
#     V1   V2   V3   V4
#1     d    b <NA>    c
#2     d    b <NA> <NA>
#3  <NA> <NA>    d    b
#4     a <NA> <NA>    b
#5     a    b    d <NA>
#6     d    c    b    c
#7     b    b    d <NA>
#8  <NA> <NA> <NA>    d
#9     a    d    d <NA>
#10 <NA>    b    d <NA>

測試數據創建代碼

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)

Answer 2

提供首選across “新” dplyr以及str_replace_all （好點@Gregor）

library(dplyr)
library(stringr)


mammographic_masses %>%
  mutate(across(everything(), 
         ~ str_replace_all(., c("na" = NA_character_, "N/A" = NA_character_))))

#>      V1   V2   V3   V4
#> 1     d    b <NA>    c
#> 2     d    b <NA> <NA>
#> 3  <NA> <NA>    d    b
#> 4     a <NA> <NA>    b
#> 5     a    b    d <NA>
#> 6     d    c    b    c
#> 7     b    b    d <NA>
#> 8  <NA> <NA> <NA>    d
#> 9     a    d    d <NA>
#> 10 <NA>    b    d <NA>

與銳相同的數據

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)

Answer 3

我們也可以使用na_if

library(dplyr)
mammographic_masses %>%
     mutate(across(everything(), ~ na_if(na_if(., "na"), "N/A")))

數據

set.seed(2020)
n <- 10
mammographic_masses <- replicate(4, sample(c(letters[1:4], "na", "N/A"), n, TRUE))
mammographic_masses <- as.data.frame(mammographic_masses)