Consider order of arguments

[hadley]

e.g.

  gen.map (
    as.data.frame,
    list( as = gen.c.of(n, gen.element(1:10) )
        , bs = gen.c.of(n, gen.element(10:20) )
        )
    )

Having the data as the last argument makes sense in (e.g.) haskell because of currying, but in R I think it makes sense to have it as the first argument, so you can use the pipe:

list(
  as = gen_element(1:10) %>% gen_c_of(n), 
  bs = gen_element(10:20) %>% gen_c_of(n)
) %>%
  gen_map(as.data.frame)

[HuwCampbell]

Makes sense. Compatibility with the ecosystem is a priority for me.

[HuwCampbell]

I have changed this up a bit, so the "simple" choice is now with (which I have changed to be flip map), I hope this gives a good choice for all.

An aside is that through these comments I discovered a bug in magrittr which slowed my working on them considerably. So I apologies if it seemed I had ignored them.