dplyr like %>% syntax in julia

Question

In R (thanks to magrittr/dplyr) you can now call functions without brackets but you can pipe them along instead.

This means that instead of coding this:

> as.character((sqrt(12)^2)
> as.Date("2014-01-01")

You could also do this:

> 12 %>% sqrt %>% .^2 %>% as.character
> "2014-01-01" %>% as.Date 

R uses this extensively to edit dataframes. Beyond dataframes, I feel that this syntax is very readable and powerful for creating functional scripts.

Does the julia language have support for something similar?

Answer 1

Yes, in two senses.

So first of all there is |>, e.g.

12 |> sqrt |> x->x^2 |> string  # 11.999999999999998
using Dates  # needed in 0.3, baked in to 0.4
"2014-01-1" |> d->Date(d,"yyyy-mm-dd") |> year |> iseven  # true

I wouldn't say its very idiomatic Julia though (or R, which the exception of doing operations on a dataframe with dplyr). There is a discussion about enhancing this type of thing and making the syntax better. You can do a lot of neat things with Lazy.jl right now though!

For DataFrames in particular, its a WIP, but there is DataFramesMeta.jl combined with Lazy.jl, which lets you do things like dplyr and LINQ such as (taken from their README):

x_thread = @> begin
    df
    @transform(y = 10 * :x)
    @where(:a .> 2)
    @by(:b, meanX = mean(:x), meanY = mean(:y))
    @orderby(:meanX)
    @select(:meanX, :meanY, var = :b)
end