Thanks--going to take a look at new package, jplyr.jl, that tshort
mentioned. Maybe what I'm looking for already exits.
On Friday, August 26, 2016 at 2:42:01 PM UTC-4, Evan Fields wrote:
>
> I'm not sure if you'll find this any cleaner, but maybe something like
>
> function summary2(df, queries...)
> d = Dict()
> for query in queries
> name = query[1]
> f = query[2]
> col = query[3]
> d[name] = f(df[col])
> end
> return d
> end
>
> Which could be used e.g.
>
> julia> summary2(dat, ("mu_x", mean, :x), ("std_y", std, :y))
> Dict{Any,Any} with 2 entries:
> "std_y" => 0.338854
> "mu_x" => 0.438158
>
> julia>
>
> Admittedly this is less flexible, but perhaps more transparent.
>