> I have a similar problem with a 1600x1600 matrix.
> I can display it on the screen, but it is very slow, as you can
> imagine.
> I'm using a formula node to adress a single matrix item. This seems to
> be much faster than to extract the item with the graphical commands of
> LabView out of the matrix and (after increasing the item by one) write
> it back.
> This should work for you as well, but it isn't fast in the sense of
> the word.
> If you have found a better solution to that problem, meanwhile, I
> would be very glad to hear about it, because this operation is the
> very heart of my programm and it slows everything down.
>
I haven't seen your code, but a common problem is to use local or global
variables inside the loop. This is inefficient
in LV since it
references the whole array. The key is to use shift registers to pass
the 2D array through your loop where you can access and replace items,
rows, etc. This will be very fast compared to using the locals and or
globals. For more details, you might want to look into the user manual
for performance tips.
Greg McKaskle