torchquad
torchquad copied to clipboard
added surface plot to visualize adaptive grid and its function values
Added function torchquad/plots/plot_adaptive_grid.py to visualize the adaptive grid with the corresponding function values on a 3D-scatter plot.
Ah, also we might want to expose the AdaptiveGrid of each integrator as well then, I think. Otherwise, users won't be able to access the grid. (we can just expose the grid through the integrator with a get_grid()
in the AdaptiveNewtonCotes integrator in the PR #135 )
I think we might also want to adjust the pointsize based on N in the grid. With N = 10 000 it's already not interpretable otherwise
Ah one more note. For high N it becomes very slow. We might consider disabling the trisurf part for N > 10000? That helps a lot with performance