Consider creating the following array of arrays:
julia> A = [rand(3) for i = 1:5]
5-element Array{Array{Float64,1},1}:
[0.007268793547957397,0.7369278377195478,0.4022138494548331]
[0.1465723997916304,0.31109494689689,0.8423556604313642]
[0.957599931474344,0.8119654625538855,0.106156261532669]
[0.8665750816190314,0.9983215743175364,0.48201520665129394]
[0.628125653002658,0.5852891135272846,0.6240444941706744]
The type of A is *Array{Array{Float64, 1}, 1}* as expected. However, let's
say I first create a 2D array and afterwards try to convert this to an
array of arrays:
julia> B = rand(3, 5)
3x5 Array{Float64,2}:
0.0941064 0.180285 0.916548 0.906796 0.309762
0.41113 0.674431 0.990499 0.577311 0.472922
0.54994 0.913673 0.898939 0.671417 0.0544504
julia> A = [B[:, i] for i = 1:5]
5-element Array{Any,1}:
[0.09410638348155964,0.4111301767644082,0.5499402666491009]
[0.1802847689705238,0.6744306262302906,0.9136732600406858]
[0.9165475568238766,0.9904993550419545,0.8989390399884916]
[0.9067959697140404,0.5773106590674384,0.6714168349642873]
[0.30976235751041514,0.4729224442912474,0.05445040118462896]
This time around the type of A is *Array{Any, 1}*. Is this to be expected?
