datatable - How to "join" with the split-apply-combine method in Julia -

i have complex join (in sql sense) perform in julia, can't figure out how working in split-apply-combine method (although can written out hand). seems should easy though. problem looks this. have dataframe of data on turtles running races:

using dataframes data = dataframe() data[:turtle] = ["suzy", "suzy", "bob", "batman", "batman", "batman", "bob"] data[:event] = ["5k", "5k", "1k", "5k", "5k", "1k", "1k"] data[:time] = [6.2 , 6.7 , 2.1, 3.2, 3.1, 0.9, 2.4] data[:photo] =["111.jpg","123.jpg","145.jpg","167.jpg","189.jpg","190.jpg","195.jpg"] data 

i datatable consists of rows of table each turtle's personal (turtlenal?) best in event ran. can need

bestfinishes = by(data, [:turtle, :event]) df      dataframe(fastesttime = minimum(df[:time])) end 

but need photo column matching rows. how do this?

well, typed in realized 1 way based on this question.

bestfinishes = by(data, [:turtle, :event]) df     dataframe(fastesttime = minimum(df[:time]), winningphoto = df[indmin(df[:time]),:photo] ) end 

however, more general way is

bestfinishes = by(data, [:turtle, :event]) df     thisfastesttime = minimum(df[:time])     df[df[:time].==thisfastesttime,:] end 

which makes things easier if want prune rows large data set efficiently. i'll see if can add example documentation since didn't seem covered (or assumed more familiarity method had).