RuntimeError: Caught an unknown exception!

[alexlim1993]

Hi all,

I have been facing this problem and couldn't resolve this issue.
the variable srips below is a SparseRipsPersistence object

Traceback (most recent call last):

  File "C:\Users\uqalim8\Documents\GitHub\topological_diagnosis\dbfil_example.py", line 50, in <module>
    pd = srips.fit_transform(db[None, :, :])

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\gtda\utils\_docs.py", line 106, in fit_transform_wrapper
    return original_fit_transform(*args, **kwargs)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\sklearn\base.py", line 699, in fit_transform
    return self.fit(X, **fit_params).transform(X)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\gtda\homology\simplicial.py", line 965, in transform
    Xt = Parallel(n_jobs=self.n_jobs)(

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\parallel.py", line 1043, in __call__
    if self.dispatch_one_batch(iterator):

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\parallel.py", line 861, in dispatch_one_batch
    self._dispatch(tasks)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\parallel.py", line 779, in _dispatch
    job = self._backend.apply_async(batch, callback=cb)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\_parallel_backends.py", line 208, in apply_async
    result = ImmediateResult(func)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\_parallel_backends.py", line 572, in __init__
    self.results = batch()

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\parallel.py", line 262, in __call__
    return [func(*args, **kwargs)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\joblib\parallel.py", line 262, in <listcomp>
    return [func(*args, **kwargs)

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\gtda\homology\simplicial.py", line 868, in _gudhi_diagram
    simplex_tree = sparse_rips_complex.create_simplex_tree(

  File "C:\Users\uqalim8\Anaconda3\lib\site-packages\gtda\externals\python\rips_complex_interface.py", line 99, in create_simplex_tree
    self.thisref.create_simplex_tree(simplex_tree.thisptr, max_dimension)

RuntimeError: Caught an unknown exception!

In the last time I encountered this issue, I managed to resolve it. It was due to having at least two same points in the dataset. So, I called np.unique or torch.unique before I feed the point cloud into fit_transform.

However, in this time, I ran into the same runtimerror despite having no same point in the dataset. For example, the following line will generate the same error.

srips.fit_transform(np.array([[[2.1234, 4.12333333], [2.1234, 4.12333334]]]))

My guess is that SparseRipsPersistence will round them up to certain percision, such that the above two points have the same value and, hence, the error.

If my guess is correct, is there any other way to avoid this problem?

[ulupo]

Sorry for the delay in responding! I haven't observed this problem before but you may well be correct in your diagnosis. I think the point is that even though the points are slightly different in their 64-bit representations, the matrix of pairwise distances (whose computation is itself subject to numerical rounding errors) might end up being identically zero.

Of course, you could always try to perturb each point's position independently with some very very small Gaussian noise. Let us know if that helps!

[alexlim1993]

That is one way to work around the problem! Thank you for the suggestion!