您可以使用 pivot_wider 创建多组交替的新列吗？

Question

我的数据目前看起来像这样，列“Number_Code 基于每个不同的 Side_Effect：

Session_ID   Side_Effect     Number_Code
 1            anxious           1
 1            dizzy             2
 1            relaxed           3
 3            dizzy             2
 7            nauseous          4
 7            anxious           1

我知道我可以做到：

mutate(rn = str_c('side_effect_', row_number())) %>% 
 pivot_wider(names_from = rn, values_from = Side_Effect)

为了创建新的列名并将每个副作用放入一个新列中，如下所示：

 session    Number_Code   side_effect1   side effect_2      side_effect_3    
      1     1                 anxious         NA                 NA
      1     2                 NA              dizzy              NA
      1     3                 NA              NA                 relaxed
      3     2                 dizzy           NA                 NA
      7     4                 nauseous        NA                 NA
      7     1                 NA              anxious            NA

但我需要根据“Side_Effect”和“Number_Code”扩大数据，并将它们放在这样的交替列中：

 session     side_effect1   number_code1   side effect_2   number_code2   side_effect_3    number_code3
        1       anxious         1              dizzy             2            relaxed          3
        3       dizzy           2               NA               NA           NA              NA
        7       nauseous        4              anxious           1            NA              NA

我看到另一篇文章，他们根据两个变量扩大了数据，但第二个的所有列都在第一个的所有列之后。有没有办法让他们像这样交替？谢谢！！

Answer 1

该pivot_wider可采取多value_from列，所以通过创建组序列后，使用pivot_wider与values_from指定感兴趣的列

library(dplyr)
library(tidyr)
df1 %>% 
   group_by(Session_ID) %>%
   mutate(rn = row_number()) %>% 
   ungroup %>% 
   pivot_wider(names_from = rn, values_from = c(Side_Effect, Number_Code))
# A tibble: 3 x 7
#  Session_ID Side_Effect_1 Side_Effect_2 Side_Effect_3 Number_Code_1 Number_Code_2 Number_Code_3
#       <int> <chr>         <chr>         <chr>                 <int>         <int>         <int>
#1          1 anxious       dizzy         relaxed                   1             2             3
#2          3 dizzy         <NA>          <NA>                      2            NA            NA
#3          7 nauseous      anxious       <NA>                      4             1            NA

如果我们需要重新排列列顺序，那么我们可以select根据数字部分和order

df1 %>% 
    group_by(Session_ID) %>%
    mutate(rn = row_number()) %>% 
    ungroup %>% 
    pivot_wider(names_from = rn, values_from = c(Side_Effect, Number_Code)) %>%
    select(Session_ID, names(.)[-1][order(readr::parse_number(names(.)[-1]))] )
# A tibble: 3 x 7
#  Session_ID Side_Effect_1 Number_Code_1 Side_Effect_2 Number_Code_2 Side_Effect_3 Number_Code_3
#       <int> <chr>                 <int> <chr>                 <int> <chr>                 <int>
#1          1 anxious                   1 dizzy                     2 relaxed                   3
#2          3 dizzy                     2 <NA>                     NA <NA>                     NA
#3          7 nauseous                  4 anxious                   1 <NA>                     NA

数据

df1 <- structure(list(Session_ID = c(1L, 1L, 1L, 3L, 7L, 7L), 
  Side_Effect = c("anxious", 
"dizzy", "relaxed", "dizzy", "nauseous", "anxious"), Number_Code = c(1L, 
2L, 3L, 2L, 4L, 1L)), class = "data.frame", row.names = c(NA, 
-6L))

Answer 2

编辑：

\n

tidyr 1.2.0使用参数可以轻松实现这一点names_vary，其中“最慢”给出交替顺序，“最快”（默认）给出块顺序。

\n

# Alternating \n\ndat |>\n  mutate(rn = row_number(), .by = Session_ID) |>\n  pivot_wider(\n    names_from = rn,\n    values_from = c(Side_Effect, Number_Code),\n    names_vary = "slowest"\n  )\n\n# A tibble: 3 \xc3\x97 7\n  Session_ID Side_Effect_1 Number_Code_1 Side_Effect_2 Number_Code_2 Side_Effect_3 Number_Code_3\n       <dbl> <chr>                 <dbl> <chr>                 <dbl> <chr>                 <dbl>\n1          1 anxious                   1 dizzy                     2 relaxed                   3\n2          3 dizzy                     2 NA                       NA NA                       NA\n3          7 nauseous                  4 anxious                   1 NA                       NA\n\n# Block\n\ndat |>\n  mutate(rn = row_number(), .by = Session_ID) |>\n  pivot_wider(\n    names_from = rn,\n    values_from = c(Side_Effect, Number_Code),\n    names_vary = "fastest"\n  )\n\n# A tibble: 3 \xc3\x97 7\n  Session_ID Side_Effect_1 Side_Effect_2 Side_Effect_3 Number_Code_1 Number_Code_2 Number_Code_3\n       <dbl> <chr>         <chr>         <chr>                 <dbl>         <dbl>         <dbl>\n1          1 anxious       dizzy         relaxed                   1             2             3\n2          3 dizzy         NA            NA                        2            NA            NA\n3          7 nauseous      anxious       NA                        4             1            NA\n

\n

第一次出现是从参数的顺序复制的values_from，因此如果需要，Number_Code出现在Side_Effect参数之前的应该是values_from = c(Number_Code, Side_Effect)。

\n