Leland McInnes comments

Results 492 comments of


                                            Leland McInnes

`Overflow encountered in true_divide error` when using Aligned UMAP

Thanks for the reproducer. I'll try to look into this when I get a little time.

UMAP enhanced clustering - Official Tutorial - ERROR: "The exit codes of the workers are {EXIT(1)}"

You shouldn't be needing anywhere near 256Gb of memory for that computation, so something is clearly going very astray. It may be a joblib related issue and not actually related...

UMAP enhanced clustering - Official Tutorial - ERROR: "The exit codes of the workers are {EXIT(1)}"

I'm afraid I don't have too much more advice I'm afraid. This seems quite peculiar.

LoweringError: Failed in nopython mode pipeline (step: native lowering)

There are a number of reasons it could be having problems. It's hard to diagnose with more detail (and the error messages from numba are unfortunately *very* long, so it...

LoweringError: Failed in nopython mode pipeline (step: native lowering)

I admit to being a little baffled. If pynndescent is installed it should be able to find NNDescent in it; is it possible envrionments are getting crossed or confused somehow?...

Does not work well on trained doc2vec model

That definitely looks underwhelming. How does t-SNE compare, or PCA? There may be less structure in the data than one might like. It looks more likely, however, that those two...

Does not work well on trained doc2vec model

If you have some time, the relevant thing to do is run the internals yourself step by step and look to see where things are getting swamped. In particular if...

Does not work well on trained doc2vec model

Hi Graham, Sorry for the very long delay on ever getting back to you on this. I got rather invested in building the new version of UMAP (which I was...

Does not work well on trained doc2vec model

Those are some nice results @vb690 ; would you mind if I referenced them in the example uses section of the documentation?

Inverse Transform + Transform does not produce correct points

Both the inverse transform and transform are stochastic approximations, so there is going to be some noise involved in this. Unfortunately this means that round-tripping like this is unlikely to...