Removing multiple variables with dplyr

Question

A data frame has the following variables:

names(df1)
"var1_a" "var1_b" "var2_a" "var2_b" "var3_a" "var4_a"

Am using dplyr to remove variables as follows - one var at a time:

df2 <- df1 %>% select(-starts_with(("var1"), everything()))
df2 <- df2 %>% select(-starts_with(("var2"), everything()))

Using "dplyr", is there a way to concatenate the two line into one?

Answer 1

You could construct a regex expression as a per your desire and use the matches wrapper in order to filter you columns accordingly (with some modifications from @BenBolker)

df2 <- df1 %>% select(-matches("^var[12]"))

Or just modify df1 in place using the %<>% pipe

library(magrittr)
df1 %<>% select(-matches("^var[12]"))