Group_by并找到组之间的相交tidyverse

Question

我正在尝试使用< code>tidyverse对< code>group_by定义的组进行< code>group_by并找到它们之间的< code>intersect。

示例数据：

< code > data = TiB ble(feature = paste 0(" Group _ "，sample(1:800，1000，replace = T))，classifier = paste0("Classifier_ "，sample(1:3，1000，replace = T)))

数据看起来像这样:

  # A tibble: 1,000 x 2
   feature   classifier  
   <chr>     <chr>       
 1 Group_629 Classifier_2
 2 Group_549 Classifier_1
 3 Group_565 Classifier_1
 4 Group_584 Classifier_2
 5 Group_181 Classifier_2
 6 Group_359 Classifier_3
 7 Group_135 Classifier_1
 8 Group_676 Classifier_3
 9 Group_731 Classifier_1
10 Group_694 Classifier_2
# … with 990 more rows

我正在尝试< code > group _ by(classifier)并找到所有组之间的< code>intersect()。

代码：

data = tibble(feature = paste0("Group_", sample(1:800, 1000, replace = T)),
              classifier = paste0("Classifier_", sample(1:3, 1000, replace = T))) %>% group_by(classifier) %>%
              mutate(intersect = intersect(feature))

产生此错误：

Error: Problem with `mutate()` input `intersect`.
x argument "y" is missing, with no default
ℹ Input `intersect` is `intersect(feature)`.
ℹ The error occurred in group 1: classifier = "Classifier_1".
Run `rlang::last_error()` to see where the error occurred.

因为我没有提供2个参数来相交。

有没有人对此有解决方案？非常感谢。

Answer 1

group_by用于数据帧上的分组计算。要获得组共有的特性的列表，您可以按分类器拆分数据并使用semi_join：

library(dplyr)

set.seed(123)

data <- tibble(feature = paste0("Group_", sample(1:800, 1000, replace = T)),
              classifier = paste0("Classifier_", sample(1:3, 1000, replace = T)))

data_split <- data %>% 
  split(.$classifier)

data_split$Classifier_1 %>% 
  semi_join(data_split$Classifier_2, by = "feature") %>% 
  semi_join(data_split$Classifier_3, by = "feature") %>% 
  pull(feature)
#>  [1] "Group_374" "Group_141" "Group_409" "Group_291" "Group_648" "Group_696"
#>  [7] "Group_55"  "Group_447" "Group_422" "Group_292" "Group_39"  "Group_358"
#> [13] "Group_177" "Group_373" "Group_464" "Group_348" "Group_468" "Group_8"  
#> [19] "Group_463" "Group_712" "Group_48"  "Group_665" "Group_229" "Group_209"
#> [25] "Group_271" "Group_229" "Group_8"   "Group_55"  "Group_55"  "Group_709"
#> [31] "Group_665" "Group_217" "Group_278" "Group_280" "Group_355" "Group_348"
#> [37] "Group_703" "Group_39"