How do I subset columns in R based on the length of the column names?

Question

I have a data table which has some columns with long, ~20 letter names such as GTEX.12WSJ.1226.SM.5LU91. I have other columns with shorter, truncated versions of those names, like GTEX.12WSJ. I need to extract the long-named columns into a separate table from the short named columns. I want both types of columns to exist on separate columns, and I want the original table to basically disappear, like rm(dt) or something.

How do I go about doing this? The closest I've come is which(nchar(names(dt)) > 9) which creates a vector of numbers going from 1 to the number of columns which fit this condition.

Answer 1

You could use split.default and split the columns based on number of characters

list_df <- split.default(df, nchar(names(df)) > 9)

where individual dataframes can be extracted with list_df[[1]] and list_df[[2]].

For example, on this dataset we divide the data into two sets. One with columns greater than 2 characters and another with less than equal to 2 characters.

df <- data.frame(abc = 1:5, ab = 1:5, a = 6:10, abcd = 11:15)
split.default(df, nchar(names(df)) > 2)

#$`FALSE`
#  ab  a
#1  1  6
#2  2  7
#3  3  8
#4  4  9
#5  5 10

#$`TRUE`
#  abc abcd
#1   1   11
#2   2   12
#3   3   13
#4   4   14
#5   5   15