Question

我有一个数据框

df1 <- data.frame(
  "player_name" = c("bob","michel","bob","michel","bob","anna"),
  "player_number" = c("1","2","1","2","1","4"),
  "position" = c("1","2","2","1","2","1"),
  "match" = c("match1","match1","match2","match2","match3","match3")
)

  player_name player_number position  match
1         bob             1        1 match1
2      michel             2        2 match1
3         bob             1        2 match2
4      michel             2        1 match2
5         bob             1        2 match3
6        anna             4        1 match3

我想把它转换成这个数据框

   match position1 position2
1 match1         1         2
2 match2         2         1
3 match3         4         1

我最多有 12 个位置，所以我可能不想硬编码
我感觉我可以使用 purr 和 group_map 来实现，但具体的功能是什么？

也许您可以将我指向 purr 中的文档和正确的功能，以便我可以先学习？（在回答之前:)

谢谢

Answer 1

您可以使用 pivot_wider：

library(tidyverse)

pivot_wider(df1, match, 
        names_from = position, values_from = player_number, names_prefix = "Position")
# A tibble: 3 x 3
  match  Position1 Position2
  <chr>  <chr>     <chr>    
1 match1 1         2        
2 match2 2         1        
3 match3 4         1

在基础 R 中你会这样做：

reshape(df1, v.names = 'player_number', 
         timevar = 'position', dir='wide', idvar = 'match', drop='player_name')

   match player_number.1 player_number.2
1 match1               1               2
3 match2               2               1
5 match3               4               1

带有data.table：

library(data.table)
setDT(df1)
dcast(df1, match~position, value.var = 'player_number')
    match 1 2
1: match1 1 2
2: match2 2 1
3: match3 4 1

如何将分组为的数据帧转换为另一个数据帧？

1 个答案: