Sparse adjoints display like dense matrices rather than sparse matrices:
julia> sprand(10,10,0.5)'
10×10 LinearAlgebra.Adjoint{Float64,SparseMatrixCSC{Float64,Int64}}:
0.653713 0.0 0.0 0.0 … 0.0 0.132038 0.994473
0.376673 0.0 0.93067 0.0 0.0 0.889498 0.11722
0.0 0.0 0.331756 0.0 0.301708 0.68326 0.0
0.491798 0.0 0.847417 0.0 0.0239453 0.0 0.66232
0.0 0.457921 0.818687 0.0 0.872932 0.705145 0.42251
0.0 0.183566 0.0 0.803265 … 0.0 0.929449 0.153552
0.0 0.0 0.34864 0.0 0.0301167 0.471377 0.277341
0.0537803 0.183686 0.731216 0.0 0.235513 0.91812 0.0
0.0 0.845597 0.200048 0.0 0.485561 0.151832 0.768555
0.0 0.985808 0.756742 0.0 0.595492 0.801145 0.356078
We need the adjoints to go through the same code path as the rest of the sparse matrix display. We aren't getting the nice terminal plots here.
Cc @dkarrasch @goggle
Most helpful comment
We need the adjoints to go through the same code path as the rest of the sparse matrix display. We aren't getting the nice terminal plots here.
Cc @dkarrasch @goggle