如何将分组变量转换为 R 中的列？

Question

I used the follwoing dataset

data.frame(Group=c("A","A","A","B","B","B"),
           time= c("10:30","10:45","10:15","10:30","10:20","10:15"),
           value= c(1,2,3,1,5,6))

and want to get the following table

I tried dcast() function from reshape2 package, but did not figure out how split Group variable based on time and value variables. Could you help?

Answer 1

library(reshape2)
dcast(df, time + value ~ paste0("group", Group), value.var = "Group")

#    time value groupA groupB
# 1 10:15     3      A   <NA>
# 2 10:15     6   <NA>      B
# 3 10:20     5   <NA>      B
# 4 10:30     1      A      B
# 5 10:45     2      A   <NA>

Answer 2

Try this:

library(dplyr)
library(tidyr)
#Code
new <- df %>%
  mutate(Val=paste0('Group',Group)) %>%
  pivot_wider(names_from = Val,values_from=Group)

Output:

# A tibble: 5 x 4
  time  value GroupA GroupB
  <chr> <dbl> <chr>  <chr> 
1 10:30     1 A      B     
2 10:45     2 A      NA    
3 10:15     3 A      NA    
4 10:20     5 NA     B     
5 10:15     6 NA     B  

Update: Using new data:

#Code 2
df %>%
  mutate(id=row_number(),
         Val=paste0('Group',Group)) %>%
  pivot_wider(names_from = Val,values_from=Group) %>% select(-id)

Output:

# A tibble: 7 x 4
  time  value GroupA GroupB
  <chr> <dbl> <chr>  <chr> 
1 10:30     1 A      NA    
2 10:30     1 A      NA    
3 10:45     2 A      NA    
4 10:15     3 A      NA    
5 10:30     1 NA     B     
6 10:20     5 NA     B     
7 10:15     6 NA     B     

Answer 3

A base R option using reshape

reshape(
  transform(
    df,
    grpName = paste0("Group", Group)
  ),
  direction = "wide",
  idvar = c("time", "value"),
  timevar = "grpName"
)

which gives

   time value Group.GroupA Group.GroupB
1 10:30     1            A            B
2 10:45     2            A         <NA>
3 10:15     3            A         <NA>
5 10:20     5         <NA>            B
6 10:15     6         <NA>            B