CODEWITHSUNDEEP
rsubsetdplyr
Michael
Michael

Reputation: 1575

subset variables in a dataset independently using dplyr

I would like to replicate the following using dplyr.

set.seed(123)
my_data <- data.frame( time = c(4:13, 4:13),
                       var1 = rep(c("a", "b"), each=10), 
                       var2 = rnorm(20))

my_data_new <- with(my_data,
                      data.frame ( time = time[var1 =="a"], 
                                  var2a = var2[var1 == "a"], 
                                  var2b = var2[var1 == "b"]))

I have tried similar syntax with transmute() and get the following error message 

Error: wrong result size (10), expected 20 or 1

Thank you!

Upvotes: 0

Views: 158

Answers (1)

akrun
akrun

Reputation: 887961

Try

library(tidyr)
res <- spread(my_data, var1, var2)
colnames(res) <- colnames(my_data_new)
identical(res, my_data_new)
#[1] TRUE

Or you can use rename from dplyr to change the column names (as suggested by @Richard Scriven)

library(dplyr)
library(tidyr)
spread(my_data, var1, var2)%>%
                        rename(var2a=a, var2b=b)

Upvotes: 5

Related Questions