r - 根据一组其他列中已识别的重复项将行折叠为多列的列表

Question

我有一个与以下虚拟数据集非常相似的大型数据集:

df = data.frame(coursecode = c("WBPH001","WBPH001","WBPH001","WBPH058","WBAS007"),
                 coursename = c("Mechanics","Mechanics","Mechanics", "Calculus 2","Introduction"),
                 courseurl = c("url1","url1","url1","url2","url3"),
                 programme_faculty = c("FSE","FSE","FSE", "FSE", "FSE"),
                 programme_name = c( "Mat","Bio","Ast","Ast","Ast"),
                 programme_ects = c("180", "180", "210", "180", "180")
                 )

这使得(所有值都是字符串):

#> print(df):
  coursecode   coursename      courseurl    programme_faculty   programme_name   programme_ects
1    WBPH001    Mechanics      url1         FSE                 Mat              180
2    WBPH001    Mechanics      url1         FSE                 Bio              180
3    WBPH001    Mechanics      url1         FSE                 Ast              210
4    WBPH058    Calculus 2     url2         FSE                 Ast              180
5    WBAS007    Introduction   url3         FSE                 Ast              180

我已经导出了整个教师的所有类(class)，但有些类(class)在多个程序中列出(在此示例中，例如与“Mat”、“Bio”和“Ast”程序关联的“Mechanics”。

简而言之，我想要实现的是删除所有这些重复的类(class)，同时保留项目信息(即名称、等、教师)。

因此，如果“coursecode”、“coursename”和“courseurl”列中存在重复项，它将自动将类(class)信息(“programme_faculty”、“programme name”和“programme_ects”列)折叠到单独的列表中每列

数据集应如下所示:

#> print(modified_df):
     coursecode coursename     courseurl    programme_faculty        programme_name      programme_ects
1    WBPH001    Mechanics      url1        c(FSE, FSE, FSE)          c(Mat, Bio, Ast)    c(180, 180, 210)
2    WBPH058    Calculus 2     url2        FSE                       Ast                 180
3    WBAS007    Introduction   url3        FSE                       Ast                 180

类(class)信息主要用于下游分析，但重要的是始终可以检索与类(class)相关的程序。因此我需要这样一个数据框，但我似乎无法找出必须使用哪些函数来实现这一点。

重要的是，字符串不能简单地折叠在一起并用“|”之类的东西分隔。

我尝试过诸如aggregate()、collapse()之类的函数，以及来自其他 stackoverflow 查询的其他建议，但他们的解决方案不适用于我的特定数据集。

Answer 1

你可以 group_by在列和 summarise这些组across要通过折叠 paste 来合并的列像这样:

library(dplyr)
df %>%
  group_by(coursecode, coursename, courseurl) %>%
  summarise(across(programme_faculty:programme_ects, ~ paste(.x, collapse = ", ")))
#> # A tibble: 3 × 6
#> # Groups:   coursecode, coursename [3]
#>   coursecode coursename   courseurl programme_faculty programme_name programme…¹
#>   <chr>      <chr>        <chr>     <chr>             <chr>          <chr>      
#> 1 WBAS007    Introduction url3      FSE               Ast            180        
#> 2 WBPH001    Mechanics    url1      FSE, FSE, FSE     Mat, Bio, Ast  180, 180, …
#> 3 WBPH058    Calculus 2   url2      FSE               Ast            180        
#> # … with abbreviated variable name ¹​programme_ects

---

您还可以list他们像这样:

library(dplyr)
df %>%
  group_by(coursecode, coursename, courseurl) %>%
  summarise(across(programme_faculty:programme_ects, ~ list(.x)))
#> # A tibble: 3 × 6
#> # Groups:   coursecode, coursename [3]
#>   coursecode coursename   courseurl programme_faculty programme_name programme…¹
#>   <chr>      <chr>        <chr>     <list>            <list>         <list>     
#> 1 WBAS007    Introduction url3      <chr [1]>         <chr [1]>      <chr [1]>  
#> 2 WBPH001    Mechanics    url1      <chr [3]>         <chr [3]>      <chr [3]>  
#> 3 WBPH058    Calculus 2   url2      <chr [1]>         <chr [1]>      <chr [1]>  
#> # … with abbreviated variable name ¹​programme_ects

^{创建于 2023 年 3 月 25 日 reprex v2.0.2}

正如@zephryl所说，你可以替换 ~list(.x)只需 list .