I’m using tonfilm’s ForceDirected (Graph) to visualize a rather big graph. Running a current test with only a fraction of the data, there’re 2x 899 IDs fed into _To ID_s and From ID of ForceDirected (Graph).
Testing a smaller set of data is all fine and dandy, but the current set of data fails to be computed by ForceDirected (Graph): A spread of 2541 values returned by From XYZ and To XYZ is all NAN.00/MaxFloat, lacking their values, the interconnections fail to be drawn. Still the text labels are where they should be. My Dual Quad-Xeon 3.0 GHz with 16GB RAM, W7, b29.2 slows down to 1 FPS. Urks.
The first 100 slices of the 2541 values spread returned by From XYZ:
The assumed final set of data will be 10 - 25x larger, I wonder if this will ever work out. What’s the current status on using vvvv on a multicore architecture? Would be DX11 beneficial in terms of performance?
Your thoughts and comments are appreciated.