Niclas Rieger

Thanks for the info! What are the number of time steps in both data sets ?

ok in that case it may be that you're indeed running into one of these edge cases mentioned above. Would you mind testing how many modes you need for each...

I'm not so sure about this anymore... Setting `variance_fraction=0.5` should result in selecting 3 and 2 principal components, which seems fine. I don't see anything unusual in the `variance_fraction` range...

hey @jiwon-j just in case you're still facing the same problem, you may give the new (xeofs version 3) [xe.cross.CCA](https://xeofs.readthedocs.io/en/latest/content/api_reference/_autosummary/xeofs.cross.CCA.html) class a try and see if it solves the issue...

Short answer: Unfortunately, no. I initially tried to implement ROCK-PCA and included it (I think prior to version 2). However, I noticed the results seemed somehow off in that I...

nice, good to know! I think a clearer error message could be really helpful for the user, maybe just mentioning that the other formats are supported.

Hi @thelightonmyway, thanks for the question! When you say the model becomes more robust with more modes, do you mean the reconstruction error decreases? If so, I agree - adding...

Oh right, I completely forgot about the rotator classes. Thanks for pointing me to `dask.graph_manipulation.wait_on` - I wasn’t aware of that one. Do you think this approach could help build...

I admit that this is a bit beyond my programming brain right now, and I can't really see how to apply `model._post_compute()` to the cross models (maybe I've just been...

Hi @n0rdp0l, thanks for the suggestion and the provided details! I wasn't familiar with supervised PCA, but I’ve briefly skimmed the papers and it does seem quite interesting. I agree...