R：pivot_wider() 的问题

Question

I have the following sample data:

varNames <- c("var1", "var2", "var3", "var4", "var5")
df <- data.frame(VariableID = rep(varNames, 5),
                 Serial = rep(1:5, 5),
                 Response = runif(25, 1, 10))

df

The form I am looking for is this:

Serial var1 var2 var3 var4 var5
1      ...  ...  ...  ...  ...
1
1
1
1
2
2
2
2
2
3
3
3
3
3
...

I was trying this:

df %>% pivot_wider(id_cols = Serial, names_from = VariableID, values_from = Response)

The output I get looks like this:

# A tibble: 5 x 6
  Serial var1      var2      var3      var4      var5     
   <int> <list>    <list>    <list>    <list>    <list>   
1      1 <dbl [5]> <NULL>    <NULL>    <NULL>    <NULL>   
2      2 <NULL>    <dbl [5]> <NULL>    <NULL>    <NULL>   
3      3 <NULL>    <NULL>    <dbl [5]> <NULL>    <NULL>   
4      4 <NULL>    <NULL>    <NULL>    <dbl [5]> <NULL>   
5      5 <NULL>    <NULL>    <NULL>    <NULL>    <dbl [5]>

What am I doing wrong? Thanks for help!

Answer 1

Try this. The issue is that rows require an id. That can be reached using row_number() . Here the code:

library(tidyverse)
#Code
newdf <- df %>% 
  arrange(Serial) %>%
  group_by(Serial) %>% mutate(id=row_number()) %>% 
  pivot_wider(names_from=VariableID,values_from=Response) %>% select(-id)

Output:

# A tibble: 25 x 6
# Groups:   Serial [5]
   Serial  var1  var2  var3  var4  var5
    <int> <dbl> <dbl> <dbl> <dbl> <dbl>
 1      1  5.52 NA       NA    NA    NA
 2      1  5.50 NA       NA    NA    NA
 3      1  5.13 NA       NA    NA    NA
 4      1  7.88 NA       NA    NA    NA
 5      1  3.49 NA       NA    NA    NA
 6      2 NA     3.68    NA    NA    NA
 7      2 NA     7.54    NA    NA    NA
 8      2 NA     9.63    NA    NA    NA
 9      2 NA     1.09    NA    NA    NA
10      2 NA     3.12    NA    NA    NA
# ... with 15 more rows

Answer 2

You may add unnest() to your function. Therefore the lists will be splittet into rows.

df %>% pivot_wider(id_cols = Serial, 
                   names_from = VariableID, 
                   values_from = Response) %>% 
       unnest()

Output:

# A tibble: 25 x 6
   Serial  var1  var2  var3  var4  var5
    <int> <dbl> <dbl> <dbl> <dbl> <dbl>
 1      1  3.51 NA       NA    NA    NA
 2      1  4.65 NA       NA    NA    NA
 3      1  4.15 NA       NA    NA    NA
 4      1  4.60 NA       NA    NA    NA
 5      1  1.74 NA       NA    NA    NA
 6      2 NA     6.42    NA    NA    NA
 7      2 NA     3.50    NA    NA    NA
 8      2 NA     1.84    NA    NA    NA
 9      2 NA     9.03    NA    NA    NA
10      2 NA     4.78    NA    NA    NA
# ... with 15 more rows

Answer 3

We can create the ids with rowid from data.table

library(dplyr)
library(data.table)
library(tidyr)
df %>% 
    arrange(Serial) %>% 
    mutate(id = rowid(Serial)) %>%
    pivot_wider(names_from = VariableID, values_from = Response) %>% 
    select(-id)

-output

# A tibble: 25 x 6
#   Serial  var1  var2  var3  var4  var5
#    <int> <dbl> <dbl> <dbl> <dbl> <dbl>
# 1      1  9.00 NA       NA    NA    NA
# 2      1  7.17 NA       NA    NA    NA
# 3      1  3.47 NA       NA    NA    NA
# 4      1  9.53 NA       NA    NA    NA
# 5      1  2.99 NA       NA    NA    NA
# 6      2 NA     1.83    NA    NA    NA
# 7      2 NA     3.65    NA    NA    NA
# 8      2 NA     5.91    NA    NA    NA
# 9      2 NA     8.91    NA    NA    NA
#10      2 NA     7.75    NA    NA    NA
# … with 15 more rows