Pedro L. C. Rodrigues
Pedro L. C. Rodrigues
Hi, Alex. Do you have any particular reason for preferring t-SNE over other manifold learning techniques? I'm asking this because lately I've been working on this tool called "Diffusion Maps"...
Ah, ok, cool then! I'll be working on it. Btw, I have a presentation here in my lab soon about the stuff I've been working on, so I made this...
I personally prefer the second option, since passing parameters to the covariance estimator could be useful for doing other stuff in the future. What you think about including a dict...
Cool. I would like to work on this! I've just read the paper and the idea seems indeed interesting.
I agree, but I'm afraid that if we start adding RPA, TLMDM and STIG in the same PR we'll get drowned, no? That said, I'm not sure how it would...
Help, I don't know how to fix this `git` problem...
I don't get why the tests fail for `flake8` at `pyriemann/utils/rotate.py:154:16: W605 invalid escape sequence '\s' ` when it does not fail, for instance, in [here](https://github.com/pyRiemann/pyRiemann/blob/aff4f500a8ca9cd430ef27652983d55da26de18c/pyriemann/utils/mean.py#L149).
Thanks @agramfort and @qbarthelemy for your inputs! I've included all your reviews and comments. You will notice now that we only have a submodule called `transfer` and that I have...
> now I think we need next to add a test_transfer.py file and start properly testing these new features. Indeed, I will propose a few tests in my next commit.
I agree and I don't know why we added this restriction to `int` values... will check it.