Gleb Minaev lounres
  • St. Petersburg, Russia – Moscow, Russia
  • Joined on 2022-09-20
lounres created pull request kscience/maps-kt#25 2024-07-08 18:11:32 +03:00
Add requirement on non-emptiness of composite trajectory parts list
lounres commented on pull request teldufalsari/kmath#1 2024-06-05 14:51:29 +03:00
Simulated annealing

Hi! The PR looks interesting. But firstly, I have organizational issues:

  1. Could you describe what you are doing? I recall what annealing is. But a brief description like "I implemented some…
lounres created pull request kscience/kmath#522 2024-05-08 22:03:22 +03:00
Fix #532 by making ShapeND a non-value class
lounres pushed to bug/defaultStridesCache at kscience/kmath 2024-05-08 22:00:08 +03:00
201887187d Make ShapeND a usual non-value class. Implement its equals and hashCode methods. Deprecate contentEquals and contentHashCode.
lounres created branch bug/defaultStridesCache in kscience/kmath 2024-05-08 22:00:07 +03:00
lounres commented on pull request kscience/kmath#521 2024-04-25 21:24:07 +03:00
WIP: feature/emd

@altavir There is a need for a function Buffer<T>.sum(elementAlgebra: Group<T>). Where should we place it?

lounres commented on pull request kscience/kmath#521 2024-04-25 21:18:05 +03:00
WIP: feature/emd

Yeah, it is better if the function is marked tailrec. But I am not sure if compiler understands the case. So I need a bit of time for a small test.

lounres commented on pull request kscience/kmath#521 2024-04-25 21:07:48 +03:00
WIP: feature/emd

Yeah, that's the idiomatic way. But place it outside the function. Or else you won't able to access it :)

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

I would recommend rewriting it with old good plain loop on indices:

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Also, similar question to this one.

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Please, move the only type argument L's bound to L's declaration cite:

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

By the way, in general you'll have to use T as L. Otherwise, either interpolate can not be resolved in snippet

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

I don't understand what the Success inheritors are for. I mean, they are all instantiated but never distinguished. All of them can be used only internally, but are cast to Success anyway.

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

You should use SeriesAlgebra's elementAlgebra in getting difference of two Double values instead of l - r. And in 8 strings below too.

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

The whole function can be rewritten in such way:

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Do not use Pair! It's non-idiomatic in Kotlin. It is really hard to always keep in mind what first and second fields hold. Instead, define and use custom data class with understandable name and understandable parameters' names.

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Could you comment what question // weird offset, is there a way to do it better? means in more detail?

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Possible typos fix suggestion:

lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd
lounres commented on pull request kscience/kmath#521 2024-04-25 18:54:03 +03:00
WIP: feature/emd

Gosh. Use when instead of long if-else sequence, please: