使用“pivot_longer()”和“names_pattern”时如何将后缀附加到“.value”

Question

我想使用pivot_longer()from {tidyr} withnames_pattern将数据转换为长格式，同时保留列名称中模式匹配之一的前缀字符串。

这似乎违反直觉，但我想在应用数据字典清理步骤之前转换为长格式，这需要原始列名。

设置

library(dplyr)
library(tidyr)

d <- tibble(id = 1,
            other_var = "foo",
            suffix_t1_value1 = "a",
            suffix_t1_value2 = "b",
            suffix_t2_value1 = "c",
            suffix_t2_value2 = "d") 

我做了什么

> pivot_longer(d,
               starts_with("suffix"),
               names_pattern = "suffix_t(1|2)_(.*)",
               names_to = c("rep", ".value"))

# A tibble: 2 x 5
     id other_var rep   value1 value2
  <dbl> <chr>     <chr> <chr>  <chr> 
1     1 foo       1     a      b     
2     1 foo       2     c      d    

所需输出

# A tibble: 2 x 5
     id other_var rep   suffix_t1_value1 suffix_t1_value2
  <dbl> <chr>     <chr> <chr>            <chr>           
1     1 foo       1     a                b               
2     1 foo       2     c                d               

我尝试过的

尝试1

> pivot_longer(d,
               starts_with("suffix"),
               names_pattern = "suffix_t(1|2)_(.*)",
               names_to = c("rep", "suffix_t1_{.value}"))

尝试2

> pivot_longer(d,
               starts_with("suffix"),
               names_pattern = "suffix_t(1|2)_(.*)",
               names_to = c("rep", paste0("suffix_t1_", ".value")))

Answer 1

我假设您想在pivot_longer. 我还没有弄清楚，如果可能的话，但如果两步过程可以，那么下面的方法应该有效：

\n\n

library(dplyr)\nlibrary(tidyr)\n\nd %>% pivot_longer(starts_with("suffix"),\n                   names_pattern = "suffix_t(1|2)_(.*)",\n                   names_to = c("rep", ".value")\n             ) %>% \n  rename_with(~ gsub("(.*)", "suffix_t1_\\\\1", .x),\n              starts_with("value"))\n\n#> # A tibble: 2 x 5\n#>      id other_var rep   suffix_t1_value1 suffix_t1_value2\n#>   <dbl> <chr>     <chr> <chr>            <chr>           \n#> 1     1 foo       1     a                b               \n#> 2     1 foo       2     c                d\n

\n

^{由reprex 包(v0.3.0)于 2021-06-09 创建}

\n\n

更新

\n

经过深入研究后pivot_longer，我认为不可能.value在内部进行访问paste，而且{.value}似乎也不支持粘合语法。

\n

然而，{tidyr} 提供了用于旋转的构建块build_longer_spec，它允许我们创建自己的my_pivot_longer函数，其中我们可以包含一个names_fn参数，该参数将函数应用于新的列名称，在这里我们可以用来gsub添加前缀或后缀。

\n\n

my_pivot_longer <- function(data,\n                            cols,\n                            names_to = "name",\n                            names_pattern = NULL,\n                            names_fn = NULL) {\n  \n  spec <- build_longer_spec(data,\n                            cols,\n                            names_pattern = names_pattern,\n                            names_to = names_to)\n\n  if (!is.null(names_fn)) {\n    fn <- rlang::as_function(names_fn)\n    spec$.value <-  fn(spec$.value)\n  }\n  \n  pivot_longer_spec(data, spec)\n    \n}\n\nd %>% \n  my_pivot_longer(starts_with("suffix"),\n                  names_pattern = "suffix_t(1|2)_(.*)",\n                  names_to = c("rep", ".value"),\n                  names_fn = ~ gsub("(.*)", "suffix_t1_\\\\1", .x))\n#> Note: Using an external vector in selections is ambiguous.\n#> \xe2\x84\xb9 Use `all_of(cols)` instead of `cols` to silence this message.\n#> \xe2\x84\xb9 See <https://tidyselect.r-lib.org/reference/faq-external-vector.html>.\n\n#> This message is displayed once per session.\n#> # A tibble: 2 x 5\n#>      id other_var rep   suffix_t1_value1 suffix_t1_value2\n#>   <dbl> <chr>     <chr> <chr>            <chr>           \n#> 1     1 foo       1     a                b               \n#> 2     1 foo       2     c                d\n

\n

^{由reprex 包(v0.3.0)于 2021-06-09 创建}

\n