r - 旋转多列 R

Question

我有这个数据:

    structure(list(classificador = c("classificador 1", "classificador 1", 
"classificador 1", "classificador 1", "classificador 1", "classificador 2", 
"classificador 2", "classificador 2", "classificador 2", "classificador 2", 
"classificador 3", "classificador 3", "classificador 3", "classificador 3", 
"classificador 3"), classe = structure(c(1L, 2L, 3L, 4L, 5L, 
1L, 2L, 3L, 4L, 5L, 1L, 2L, 3L, 4L, 5L), .Label = c("bolha", 
"coral_sol", "cosc_mult", "coscinodiscus", "detritus", "fake_coral", 
"multiples", "org_parts", "organism", "shadow"), class = "factor"), 
    media_entrou = c(0.000181463342477325, 9.11739514644428e-06, 
    0, 0.00710836752524651, 0.0523307152250179, 5.76207054303695e-05, 
    2.31948309497038e-05, 0.000533722409264464, 0.00395244621833009, 
    0.0277347189726644, 0.000511091116817504, 2.78792287970657e-05, 
    0.00167363838758039, 0.00714582937886204, 0.03664291179572
    ), desv_pad_entrou = c(0.000625138593404583, 4.01259397230609e-05, 
    0, 0.00982865377212798, 0.0382102106478645, 0.000197304670837601, 
    0.00015803336539453, 0.00133833325218429, 0.00460178168250652, 
    0.0198204066019173, 0.00560789561914724, 0.000295803787791627, 
    0.0011439791450099, 0.00660048491487252, 0.023022755540947
    ), media_saiu = c(6.31000967592145e-05, 0, 0, 0.0117956025289566, 
    0.00890647892432332, 0.000100059064539586, 0, 0, 0.0587231357033222, 
    0.0639343116405082, 0.00028538070772188, 0, 2.54448072255982e-05, 
    0.0616763969528522, 0.0284492872735916), desv_pad_saiu = c(0.00010828405298676, 
    0, 0, 0.00782886688427178, 0.00460945360308487, 0.000418014988563, 
    0, 0, 0.0258394941887122, 0.0678523075381303, 0.000824409079665752, 
    0, 0.000147663248976053, 0.0346114091736888, 0.0165290142815091
    ), N_ciclos = c(118L, 118L, 118L, 118L, 118L, 318L, 318L, 
    318L, 318L, 318L, 158L, 158L, 158L, 158L, 158L)), class = c("grouped_df", 
"tbl_df", "tbl", "data.frame"), row.names = c(NA, -15L), groups = structure(list(
    classificador = c("classificador 1", "classificador 2", "classificador 3"
    ), .rows = structure(list(1:5, 6:10, 11:15), ptype = integer(0), class = c("vctrs_list_of", 
    "vctrs_vctr", "list"))), class = c("tbl_df", "tbl", "data.frame"
), row.names = c(NA, -3L), .drop = TRUE))

我想用新列对数据进行透视:classificador、classe、E_S、media、和 desvio。其中 E_S 列的值名称 entrou 对应于列 media_entrou 和 desvio_entrou 以及值名称 saiu 根据 media_saiu 和 desvio_saiu 列以及 media_entrou 和 media_saiu 列的值新列 media 以及新列 desvio 中列 desvio_entrou 和 desvio_saiu 的值。

前 4 行的预期输出:

<表类="s-表"><头>分类器类E_S媒体设计N_ciclos<正文>分类器 1波哈进入0.0001810.000625118分类器 1波哈再见0.00006310.000108118分类器 1珊瑚溶胶进入0.000009120.0000401118分类器 1珊瑚溶胶再见00118

谢谢

Answer 1

我们可以使用 pivot_longer，指定正则表达式模式来捕获 names_pattern 中的组，即 ^([^_]+) - 匹配一个或多个不是 _ 的字符，从一开始 (^) 和捕获 ((...)) 后跟 _ 并从列名中捕获其余字符 (.*)，这些字符将成为“E_S”列中的值

library(tidyr)
library(dplyr)
library(stringr)
df1 %>% 
   ungroup %>% 
   select(-N_ciclos) %>%
   dplyr::rename_with(~ str_remove(., "_pad")) %>% 
   pivot_longer(cols = contains("_"), names_to = c(".value", "E_S"),
      names_pattern = "^([^_]+)_(.*)")

-输出

# A tibble: 30 × 5
   classificador   classe        E_S         media      desv
   <chr>           <fct>         <chr>       <dbl>     <dbl>
 1 classificador 1 bolha         entrou 0.000181   0.000625 
 2 classificador 1 bolha         saiu   0.0000631  0.000108 
 3 classificador 1 coral_sol     entrou 0.00000912 0.0000401
 4 classificador 1 coral_sol     saiu   0          0        
 5 classificador 1 cosc_mult     entrou 0          0        
 6 classificador 1 cosc_mult     saiu   0          0        
 7 classificador 1 coscinodiscus entrou 0.00711    0.00983  
 8 classificador 1 coscinodiscus saiu   0.0118     0.00783  
 9 classificador 1 detritus      entrou 0.0523     0.0382   
10 classificador 1 detritus      saiu   0.00891    0.00461  
# … with 20 more rows