GROUP_CONCAT与dplyr或R [英] GROUP_CONCAT with dplyr or R

问题描述

我很难在dplyr中复制典型SQL GROUP_CONCAT函数的功能.我还要确保可以控制组内的顺序.理想情况下，我想使用hadleyverse/tidyverse，但是base R或其他软件包也可以使用.

I am having difficulty replicating the functionality of a typical SQL GROUP_CONCAT function in dplyr. I would also like to make sure the ordering inside the groups can be controlled. Ideally I want to use the hadleyverse/tidyverse but base R or other packages will work too.

示例数据:

ID    name
1     apple
1     orange
2     orange
3     orange
3     apple

所需的输出:

ID    name
1     apple,orange
2     orange
3     apple,orange

请注意，对于ID = 3，顺序是字母顺序，而不是行的顺序.我认为可以先进行 arrange 处理，但是最好在 summarise 语句之内进行控制.

Note that for ID=3, the ordering is in alpha order, not how the rows are ordered. I think this can probably be handled by doing an arrange first, but it would be nice to control inside the summarise statement or the like.

推荐答案

在 R 中，我们可以使用group by操作之一.

In R, we can use one of the group by operations.

library(dplyr)
df1 %>%
    group_by(ID) %>%
    summarise(name = toString(sort(unique(name))))
#     ID          name 
#   <int>         <chr>
#1     1 apple, orange
#2     2        orange
#3     3 apple, orange

---

或使用 data.table

library(data.table)
setDT(df1)[, .(name = toString(sort(unique(name)))), by = ID]
#   ID          name
#1:  1 apple, orange
#2:  2        orange
#3:  3 apple, orange

这篇关于GROUP_CONCAT与dplyr或R的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！