R 折叠行(如果在同一组中)
大家好,我有一个数据帧,例如:R 折叠行(如果在同一组中),r,dplyr,R,Dplyr,大家好,我有一个数据帧,例如: Group family 1 A Canidae 2 B Canidae 3 A Felidae 4 B Canidae 5 C Elephantidae 6 C Galinacae 7 D Galinacae 8 D Siuridae 9 E Apidae 我想折叠组存在家族的组(例如: 犬科存在于A和
Group family
1 A Canidae
2 B Canidae
3 A Felidae
4 B Canidae
5 C Elephantidae
6 C Galinacae
7 D Galinacae
8 D Siuridae
9 E Apidae
我想折叠组
存在家族
的组(例如:
犬科存在于A和B
因此,我将每个组的所有唯一值合并并添加到family2
中
Group family2
A,B Canidae,Felidae
然后我继续,我看到象甲科和加利那科都在C
中,thta加利那科也在D
中,所以我崩溃了:
Group family2
A,B Canidae,Felidae
C,D Elephantidae,Galinacae,Siuridae
最后,我们应该得到:
Group family2
A,B Canidae,Felidae
C,D Elephantidae,Galinacae,Siuridae
E Apidae
有人有主意吗
这里是为了一个像这样的东西的数据请?非常感谢您的帮助和时间
以下是数据(如果有帮助):
structure(list(Group = structure(c(1L, 2L, 1L, 2L, 3L, 3L, 4L,
4L, 5L), .Label = c("A", "B", "C", "D", "E"), class = "factor"),
family = structure(c(2L, 2L, 4L, 2L, 3L, 5L, 5L, 6L, 1L), .Label = c("Apidae",
"Canidae", "Elephantidae", "Felidae", "Galinacae", "Siuridae"
), class = "factor")), class = "data.frame", row.names = c(NA,
-9L))
这是我的解决方案,其中包含一些查找函数
# A lookup function that look for intersect between group
# if there are at least one intersect - those group will be combined
look_up_group <- function(one_group, lookup_list) {
matched_list <- map(lookup_list, function(x) {
tryCatch(
{
intersect(x, one_group)
}, error = function(e) {
stop(paste0("Error in lookup function: one_group=", one_group, "; x=", x))
})
})
index <- which(unlist(map(matched_list, function(x) { length(x) > 0 })))
sort(unique(unlist(lookup_list[index])))
}
df %>%
# First remove all duplicated rows - exactly the same for both Group, Family
filter(!duplicated(.)) %>%
# arrange in alphabetical order
arrange(Group, family) %>%
# create a Group_2 which is combination of all Group for each family
group_by(family) %>%
mutate(Group_2 = list(Group)) %>%
ungroup() %>%
# Create Group_3 which is the full combined Group for all intersect Group
mutate(Group_3 = map(.[["Group_2"]], function(x) { look_up_group(one_group = x, lookup_list = .[["Group_2"]]) })) %>%
# Combine all Group_3 into a Group_final
mutate(Group_final = unlist(map(Group_3, function(x) { paste (x, collapse = ",")} ))) %>%
# Finally put them all together.
select(Group_final, family) %>%
group_by(Group_final) %>%
summarize(family = paste(family, collapse = ","), .groups = "drop")
然后下一步创建组_3并将它们合并到组_final中
最终结果
# A tibble: 3 x 2
Group_final family
* <chr> <chr>
1 A,B Canidae,Felidae,Canidae
2 C,D Elephantidae,Galinacae,Galinacae,Siuridae
3 E Apidae
#一个tible:3 x 2
最后一组
*
1 A,B犬科,猫科,犬科
2 C,D象皮动物科,象皮动物科,象皮动物科,象皮动物科,象皮动物科
3.蜜蜂科
[更新:为调试添加了tryCatch]这感觉很熟悉-您最近是否发布(并删除)了类似的问题?;)请参阅和<代码>g=数据帧(d)中的图形<代码>d$m=组件(g)$membership[d$Group];然后按“m”粘贴折叠非常感谢,它确实起作用了,因为实际上我收到了以下错误消息:“``问题与
mutate()
inputGroup_3
。x无法将类型“closure”强制为类型“character”的向量ℹ 输入Group_3
为map(…)
``你知道发生了什么事吗?@chippycentra-不确定没有实际数据-但很明显,你的数据中有一种情况导致函数intersect出现错误
-我添加了一个tryCatch,只是为了打印出导致错误的函数-你可以更新函数代码并运行它,看看是什么导致了错误吗?
# remove duplicate & create variable Group_2
tmp <- df %>%
filter(!duplicated(.)) %>%
arrange(Group, family) %>%
group_by(family) %>%
mutate(Group_2 = list(Group)) %>%
ungroup()
Group family Group_2
<fct> <fct> <list>
1 A Canidae <fct [2]>
2 A Felidae <fct [1]>
3 B Canidae <fct [2]>
4 C Elephantidae <fct [1]>
5 C Galinacae <fct [2]>
6 D Galinacae <fct [2]>
7 D Siuridae <fct [1]>
8 E Apidae <fct [1]>
> tmp$Group_2
[[1]]
[1] A B
Levels: A B C D E
[[2]]
[1] A
Levels: A B C D E
[[3]]
[1] A B
Levels: A B C D E
[[4]]
[1] C
Levels: A B C D E
[[5]]
[1] C D
Levels: A B C D E
[[6]]
[1] C D
Levels: A B C D E
[[7]]
[1] D
Levels: A B C D E
[[8]]
[1] E
Levels: A B C D E
# Create Group_3
tmp <- tmp %>%
mutate(Group_3 = map(.[["Group_2"]],
function(x) { look_up_group(one_group = x, lookup_list = .[["Group_2"]]) })) %>%
mutate(Group_final = unlist(map(Group_3, function(x) { paste (x, collapse = ",")} )))
# A tibble: 8 x 5
Group family Group_2 Group_3 Group_final
<fct> <fct> <list> <list> <chr>
1 A Canidae <fct [2]> <fct [2]> A,B
2 A Felidae <fct [1]> <fct [2]> A,B
3 B Canidae <fct [2]> <fct [2]> A,B
4 C Elephantidae <fct [1]> <fct [2]> C,D
5 C Galinacae <fct [2]> <fct [2]> C,D
6 D Galinacae <fct [2]> <fct [2]> C,D
7 D Siuridae <fct [1]> <fct [2]> C,D
8 E Apidae <fct [1]> <fct [1]> E
tmp %>%
select(Group_final, family) %>%
group_by(Group_final) %>%
summarize(family = paste(family, collapse = ","), .groups = "drop")
# A tibble: 3 x 2
Group_final family
* <chr> <chr>
1 A,B Canidae,Felidae,Canidae
2 C,D Elephantidae,Galinacae,Galinacae,Siuridae
3 E Apidae