Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

What to do with matlab error: max iterations per pass excedded? #3

Open
nhemleben opened this issue Jul 12, 2019 · 4 comments
Open

Comments

@nhemleben
Copy link
Contributor

While running isosplit5_mex() I get the above warning, sometimes multiple times when running on a single data set. I tried running isosplit5() with max_iterations = 2^30 but the warning persists.
I do not understand what the algorithm does once it breaks from the iterations loop. Does it return the sub-optimal labels, or does it keep going after reassigning something?

I am curious if you think this warning is worth being concerned about and rerunning the clustering.

@magland
Copy link
Collaborator

magland commented Sep 14, 2019

Not sure.... if the data is reasonable size, could you share it? Then I will test.

@nhemleben
Copy link
Contributor Author

The data here should produce the error when running the algorithm on the full space.

@magland
Copy link
Collaborator

magland commented Oct 28, 2019

the data are 3.5 GB -- pretty large for a basic test. I am guessing that there are duplicate points -- meaning more than one data point with the same coordinates.

@magland
Copy link
Collaborator

magland commented Oct 28, 2019

Hi @nhemleben. Just to clarify... is this a warning or an error? (i.e., does it prevent the clustering from finishing)?

If you are able to run this in Python, it would be helpful to see if you get the same problem with the Python-wrapped version (which is available on PyPI). See: https://github.com/magland/isosplit5_python

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants