Delete missing values after the last non missing value in each id

[rafael.guerra]

Check this option:

vcat([g[1:findlast(!isequal(missing), g.b),:] for g in groupby(df,:id)]...)