DataFrames: conditional probabilities

General Usage
5

This seems better:

julia> @chain dfYX begin
           innerjoin(dfZY; on=:y)
           groupby([:z,:x])
           combine([:probYX, :probZY] => ((yx,zy)->sum(yx.*zy)) => :probZX)
       end
4×3 DataFrame
 Row │ z      x      probZX  
     │ Int64  Int64  Float64 
─────┼───────────────────────
   1 │     1      1     0.43
   2 │     1      2     0.31
   3 │     2      1     0.57
   4 │     2      2     0.69
2 Likes