max() on sparse matrices

[nassarhuda]

max() on 2 sparse matrices returns an Array{Float64,Int64}, can this be changed to return a sparse matrix as well?

julia> A = sprand(5,5,1/5)
5x5 sparse matrix with 2 Float64 entries:
    [4, 2]  =  0.502774
    [5, 2]  =  0.224851

julia> B = sprand(5,5,1/5)
5x5 sparse matrix with 6 Float64 entries:
    [1, 1]  =  0.359682
    [5, 1]  =  0.608841
    [4, 2]  =  0.989943
    [2, 3]  =  0.671907
    [5, 4]  =  0.34389
    [4, 5]  =  0.0793254

julia> C = max(A,B)
5x5 Array{Float64,2}:
 0.359682  0.0       0.0       0.0      0.0      
 0.0       0.0       0.671907  0.0      0.0      
 0.0       0.0       0.0       0.0      0.0      
 0.0       0.989943  0.0       0.0      0.0793254
 0.608841  0.224851  0.0       0.34389  0.0    

[tkelman]

Makes sense to me. Best to post a cross-reference link when you post both an issue and a related mailing list thread, here https://groups.google.com/forum/#!topic/julia-users/Qh-7q0U5etA

Looks like @which max(A,B) is falling back to the AbstractArray definition from @vectorize_2arg. Could probably get away with adding max and min to the specialized sparse broadcast loop here.