speed up in update method

[Sourmpis]

Hi! Thanks for sharing the library!
I noticed that if you change the loop in the update method with an einsum operation you can speed up the training by some amount.
Hope you find it useful.
Christos

[JustGlowing]

Thanks for your contribution @Sourmpis !

I'm also working on speeding up minisom, but I'm more looking into using numba. What sort of speedup did you achieve with einsum?

[Sourmpis]

Happy if this is helpful! Numba sounds really good.
Using einsum, basically you avoid doing the multiplication of g neighborhood matrix and the matrix x_w = x - self._weights in a loop but instead you doing it in numpy.

[JustGlowing]

Of course, did you measure the speed up? Is it 2x, 4x, ... ?

[Sourmpis]

sorry i misunderstood the question, the speedup is around x2.5

[JustGlowing]

That is amazing!

Can you fix the following codestyle issues?

$ pycodestyle minisom.py 
minisom.py:5:80: E501 line too long (84 > 79 characters)
minisom.py:249:5: E303 too many blank lines (3)

[Sourmpis]

just did it :)

[JustGlowing]

Thanks @Sourmpis ! I'll now merge this and you'll be mentioned in the release notes of the new version of minisom ;-)

[Sourmpis]

Thank you!

[JustGlowing]

hi there, I'm thinking about changing the license of Minisom to MIT (or GPL). This will allow Minisom to have a paper on the Journal of Open Source software. This WILL NOT affect the ownership of your contribution and does not imply that I will profit from Minisom. Your contribution was very welcome and I'd like to have your approval.

[Sourmpis]

Sure no problem :)

…

On Thu, Jul 23, 2020 at 6:46 PM Giuseppe Vettigli ***@***.***> wrote: hi there, I'm thinking about changing the license of Minisom to MIT (or GPL). This will allow Minisom to have a paper on the Journal of Open Source software. This WILL NOT affect the ownership of your contribution and does not imply that I will profit from Minisom. Your contribution was very welcome and I'd like to have your approval. — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#33 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AGSFNHRIMVOJ4CST5J6PA5LR5BSO5ANCNFSM4HQEZXPQ> .