More benchmark ideas

[lukashaenjes]

• filter by columns: base df[df$x == df$y, ] vs. df %>% dplyr::filter(x == y) vs. dplyr::filter(df, x == y) (i.e., no pipe) vs. existing data.table solutions
• join: base::merge() vs. data.table::merge() vs. dplyr::*_join()

[sw-jakobgepp]

• x[which(x in y)] is intersect(x, y)