Subset with regex

[arnold-c]

I was wondering if it's possible (or possible to add as a feature) to use @rsubset with regex? It would be incredibly helpful for my workflow where I often stack DataFrames with similar variable names i.e. they share characters and I want to subset them all based on a pattern. I appreciate that it is possible to do this with @select before stacking, but there are times where it would be useful to apply this directly to the long dataframe.

MWE below.

using DataFrames, DataFramesMeta

wide_df = DataFrame(
    S_L = rand(100),
    S_M = rand(100),
    S_H = rand(100),
    I_L = rand(100),
    I_M = rand(100),
    I_H = rand(100),
    R_L = rand(100),
    R_M = rand(100),
    R_H = rand(100)
)

long_df = stack(wide_df, variable_name = :state, value_name = :proportion)

let
    y = $(r"R")       #or y = Set([$(r"R")])
    @rsubset long_df :state in y         #or :state == $(r"R")
end

[bkamins]

@rsubset long_df contains(:state, r"R")

[arnold-c]

That's very helpful. Thanks @bkamins!