pytorch-topological
pytorch-topological copied to clipboard
Persistence diagram and wasserstein distance using ripser vs torch_topological
Hello,
- I have used ripser and torch_topological libraries to compute cubical persistence of the same image but the two two diagrams are different.
- About the wasserstein distance, I have tried to input same persistence diagram to persim.wasserstein and torch_topological.nn.wasserstein distance with same setting but the answers are different.
Can you please guide me what is the reason behind this? Am I doing something wrong? Is there a way to resolve this issue?
Ripser/Persim "diagram1 = np.array([[0., 1.]]) diagram2 = np.array([[0., 1.], [0., 1.]]) distance = persim.wasserstein(diagram2, diagram1) output: Wasserstein distance: 0.7071067811865476"
with torch_topological I got "tensor(0.5000)"
Thanks in advance.