R data.table: subsetting data.table/dataframe 基于行值的大小

Question

这是一个基本问题，但我被难住了：

我有以下 R data.table:

library(data.table)
DT <- fread('unique_point biased    data_points   team   groupID                                                                                                           
 up1          FALSE     3             A      xy28352                                                                                                                 
 up1          TRUE      4             A      xy28352                                                                                                                 
 up2          FALSE     1             A      xy28352                                                                                                                  
 up2          TRUE      0             X      xy28352                                                                                                                  
 up3          FALSE     12            Y      xy28352                                                                                                                 
 up3          TRUE      35            Z      xy28352')

打印为

> DT
   unique_point biased data_points team groupID
1:          up1  FALSE           3    A xy28352
2:          up1   TRUE           4    A xy28352
3:          up2  FALSE           1    A xy28352
4:          up2   TRUE           0    X xy28352
5:          up3  FALSE          12    Y xy28352
6:          up3   TRUE          35    Z xy28352

team 列的值是字母 A 到 Z，26 种可能性。在这一刻。如果我用这段代码计算行值：

DT[, counts := .N, by=c("team")]

这给出了

> DT
   unique_point biased data_points team groupID counts
1:          up1  FALSE           3    A xy28352      3
2:          up1   TRUE           4    A xy28352      3
3:          up2  FALSE           1    A xy28352      3
4:          up2   TRUE           0    X xy28352      1
5:          up3  FALSE          12    Y xy28352      1
6:          up3   TRUE          35    Z xy28352      1

我想在 DT 中创建 26 个新列，它给出每个 team、A、B、C 等的大小.

结果 data.table 看起来像：

> DT
   unique_point biased data_points team groupID    A   B   C ... Z
1:          up1  FALSE           3    A xy28352    3   0   0 ... 1
2:          up1   TRUE           4    A xy28352    3   0   0 ... 1
3:          up2  FALSE           1    A xy28352    3   0   0 ... 1
4:          up2   TRUE           0    X xy28352    3   0   0 ... 1
5:          up3  FALSE          12    Y xy28352    3   0   0 ... 1
6:          up3   TRUE          35    Z xy28352    3   0   0 ... 1

我不确定如何使用 data.table 语法做到这一点..

编辑：我也很高兴用 base R 和 dplyr 来做这件事。

Answer 1

那plyr呢，可以吗？

library(data.table)
library(plyr)

DT <- fread('unique_point biased    data_points   team   groupID                                                                                                           
            up1          FALSE     3             A      xy28352                                                                                                                 
            up1          TRUE      4             A      xy28352                                                                                                                 
            up2          FALSE     1             A      xy28352                                                                                                                  
            up2          TRUE      0             X      xy28352                                                                                                                  
            up3          FALSE     12            Y      xy28352                                                                                                                 
            up3          TRUE      35            Z      xy28352')

ldply(LETTERS, function(x){
  n <- nrow(DT[team == as.character(x),])
  DT[, as.character(x) := n]
  return(DT[team == x,])
})

> DT
   unique_point biased data_points team groupID A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
1:          up1  FALSE           3    A xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
2:          up1   TRUE           4    A xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
3:          up2  FALSE           1    A xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
4:          up2   TRUE           0    X xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
5:          up3  FALSE          12    Y xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1
6:          up3   TRUE          35    Z xy28352 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1

Answer 2

这是一个不寻常的解决方案，但它确实有效。我使用了 dplyr 和 tidyr

DT[, counts := .N, by=c("team")]
x <- data.frame(team = sample(LETTERS,26))%>%arrange(team)
y <- DT%>%select(team,counts)%>%unique()
df <- x%>%left_join(y,"team")%>%spread(team, counts,fill = 0)
cbind(DT,df)

注意：left_join 确实会发出警告消息，但不会篡改输出，并且有解决方法

R data.table: subsetting data.table/dataframe 基于行值的大小

R data.table: subsetting data.table/dataframe based on size of row value

r

dataframe

melt

unfold

data.table