Extract multiple variables by naming convention, for more than two types of naming convention - r

Question

I'm trying to extract multiple variables that start with certain strings. For this example I'd like to write a code that will extract all variables that start with X1 and Y2.
set.seed(123)
df <- data.frame(X1_1=sample(1:5,10,TRUE),
X1_2=sample(1:5,10,TRUE),
X2_1=sample(1:5,10,TRUE),
X2_2=sample(1:5,10,TRUE),
Y1_1=sample(1:5,10,TRUE),
Y1_2=sample(1:5,10,TRUE),
Y2_1=sample(1:5,10,TRUE),
Y2_2=sample(1:5,10,TRUE))
I know I can use the following to extract variables that begin with "X1"
Vars_to_extract <- c("X1")
tempdf <- df[ , grep( paste0(Vars_to_extract,".*" ) , names(df), value=TRUE)]
X1_1 X1_2
1 3 5
2 3 4
3 2 1
4 2 2
5 3 3
But I need to adapt above code to extract variables multiple variable types, if specified like this
Vars_to_extract <- c("X1","Y2")
I've been trying to do it using an %in% with .* within the grep part, but with little success. I know to I can write the following which is pretty manual, merging each set of variables separately.
tempdf <- data.frame(df[, grep("X1.*", names(df), value=TRUE)] , df[, grep("Y2.*", names(df), value=TRUE)] )
X1_1 X1_2 Y2_1 Y2_2
1 3 5 1 5
2 3 4 1 5
3 2 1 2 3
4 2 2 3 1
5 3 3 4 2
However, in real world situation, I often work with lots of variables and would have to do this numerous times. Is it possible to write it in this way using %in% or does I need use a loop? Any help or tips will be gratefully appreciated. Thanks

Answer 1

We could use contains if we want to extract column names that have the substring anywhere in the string
library(dplyr)
df %>%
select(contains(Vars_to_extract))
Or with matches, we can use a regex to specify the the string starts (^) with the specific substring
library(stringr)
df %>%
select(matches(str_c('^(', Vars_to_extract, ')', collapse="|")))
With grep, we could create a single pattern by paste with collapse = "|"
df[grep(paste0("^(",paste(Vars_to_extract, collapse='|'), ")"), names(df))]
# X1_1 X1_2 Y2_1 Y2_2
#1 3 5 5 3
#2 3 3 5 5
#3 2 3 3 3
#4 2 1 1 2
#5 3 4 4 5
#6 5 1 1 5
#7 4 1 1 3
#8 1 5 3 2
#9 2 3 4 2
#10 3 2 1 2
Or another approach is to startsWith with lapply and Reduce
df[Reduce(`|`, lapply(Vars_to_extract, startsWith, x = names(df)))]