数据分析：筛选多组交集特征

介绍

有时候需要在多个组间筛选它们的交集特征，本文利用R语言实现该目的

加载R包

library(UpSetR)
library(tidyverse)
1
2

Upset画图

movies <- read.csv(system.file("extdata", "movies.csv", package = "UpSetR"), 
                   header = T, sep = ";")
movies_list <- list(
  Action = movies %>%
    dplyr::filter(Action == 1) %>%
    dplyr::pull(Name),
  Adventure = movies %>%
    dplyr::filter(Adventure == 1) %>%
    dplyr::pull(Name),
  Children = movies %>%
    dplyr::filter(Children == 1) %>%
    dplyr::pull(Name),
  Comedy = movies %>%
    dplyr::filter(Comedy == 1) %>%
    dplyr::pull(Name),
  Crime = movies %>%
    dplyr::filter(Crime == 1) %>%
    dplyr::pull(Name),
  Documentary = movies %>%
    dplyr::filter(Documentary == 1) %>%
    dplyr::pull(Name)  
)

movies_pl <- UpSetR::upset(
  data = fromList(movies_list),
  nsets = 3, 
  sets = c("Action", "Adventure", "Children", 
           "Comedy", "Crime", "Documentary"),
  order.by = "freq",
  main.bar.color = "gray10",
  sets.bar.color = "gray",
  matrix.color = "gray10",
  mainbar.y.label = "NO. of movies",
  sets.x.label = "NO. of movies")

movies_pl
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36

判断交集特征

• 去冗余变量 df_uniq_movie

• 分组变量标签 df_group_movie

df_uniq_movie <- data.frame(feature = unique(unlist(movies_list)))
df_group_movie <- lapply(movies_list, function(x){
  data.frame(feature = x)
}) %>% 
  dplyr::bind_rows(.id = "Sequence")
1
2
3
4
5
6

• 给变量打上交集标签

df_int_movie <- lapply(df_uniq_movie$feature, function(x){
  intersection <- df_group_movie %>% 
    dplyr::filter(feature == x) %>% 
    dplyr::arrange(Sequence) %>% 
    dplyr::pull(Sequence) %>% 
    paste0(collapse = "|")
  # build the dataframe
  return(data.frame(feature = x, int = intersection))
}) %>% 
  dplyr::bind_rows()

head(df_int_movie)
1
2
3
4
5
6
7
8
9
10
11
12