I have this data frame,
set.seed(123)
dataf <- data.frame( id = seq(1:5),
A_1 = rnorm(5),
A_2 = rnorm(5),
A_3 = rnorm(5),
A_4 = rnorm(5),
A_5 = rnorm(5))
> dataf
id A_1 A_2 A_3 A_4 A_5
1 1 -0.56047565 1.7150650 1.2240818 1.7869131 -1.0678237
2 2 -0.23017749 0.4609162 0.3598138 0.4978505 -0.2179749
3 3 1.55870831 -1.2650612 0.4007715 -1.9666172 -1.0260044
4 4 0.07050839 -0.6868529 0.1106827 0.7013559 -0.7288912
5 5 0.12928774 -0.4456620 -0.5558411 -0.4727914 -0.6250393
And I have a vector like the following, based on which, I want to rename the above data frame variables.
sep <- c(0, 1, 1, 0, 0)
In this vector, I have the marker based on which the varibles should be renamed. If the marker says 0, no change. If the marker say 1, we add a prefix to the variable name, say, _mod. For example, the second element of sep is 1, that means I need to rename the second variable A_2 to A_2_mod. The resulting table should look like this:
id A_1 A_2_mod A_3_mod A_4 A_5
1 1 -0.56047565 1.7150650 1.2240818 1.7869131 -1.0678237
2 2 -0.23017749 0.4609162 0.3598138 0.4978505 -0.2179749
3 3 1.55870831 -1.2650612 0.4007715 -1.9666172 -1.0260044
4 4 0.07050839 -0.6868529 0.1106827 0.7013559 -0.7288912
5 5 0.12928774 -0.4456620 -0.5558411 -0.4727914 -0.6250393
I can do it easily by renaming by position. However, I am wondering how to make the renaming scalable in case, say, I have thousands of columns.
You can use dplyr::rename_with to help. For example
dplyr::rename_with(dataf,
function(x) paste0(x, "_mod"),
.cols=num_range("A_", which(sep==1))
)
I used the num_range() tidyselect helper to choose just the A columns that have a 1 in the corresponding position in sep