Edge Rewrite
// HTMLRewriter · presentation

This page was redesigned at the edge.

Cloudflare fetched the original article and streamed it through HTMLRewriter to apply an entirely new visual system without rebuilding the source page.

// request.cf · coarse context

A page that knows where it met you.

Only coarse request metadata is shown. This demo does not display or persist visitor IP addresses.

Country
US
Cloudflare location
CMH
Connection
HTTP/2
Language
Not provided

Ray ID: a22581d3cc0dc465

Jump to content

Talk:Comparison of Gaussian process software

Page contents not supported in other languages.
Add topic
From Wikipedia, the free encyclopedia
Latest comment: 1 year ago by Joanico in topic Delete CODES

Disclosure

[edit]

I'm the author of a gaussian process software, but I did not list it here. Joanico (talk) 22:22, 23 May 2020 (UTC)Reply

Linear transformations in GPVecchia

[edit]

From the vignette https://github.com/katzfuss-group/GPvecchia/blob/master/vignettes/GPvecchia_vignette.Rmd (the only documentation I could find about GPVecchia) it appears that linear transformations can be applied to predictions, but not to the process before fitting the data, the latter being the intended meaning of the columns, explained above in the page, and motivated by the fact that the former is simpler and can be done independently of the specific program after fetching the predictions.

So I put a "Yes?" in gray instead of a clear green "Yes".

Joanico (talk) 19:21, 1 September 2021 (UTC)Reply

Reading the reference manual was sufficient to determine that 1) indeed linear transformations are not supported, and 2) neither correlated errors.

Joanico (talk) 21:05, 1 September 2021 (UTC)Reply

Delete MatlabGP

[edit]

I deleted the recently added MatlabGP software because on a quick skim I have found evidence of low competence, namely this line of code. I won't say why that line is bad in case the author tried to "guess the teacher password" by changing that line to what I say. If you are the author and think this is unfair, please tell me your guess about why I am thinking that line is not ok, and explain why it's actually ok instead. Joanico (talk) 13:39, 28 September 2024 (UTC)Reply

Delete SAMBO

[edit]

I looked at the code on github (a85f8fb). Its implementation of GP regression is a thin wrapper around scikit-learn's one. Multiple values in the table look wrong. It seems to have been developed in the last 2-3 months. Joanico (talk) 21:33, 2 March 2025 (UTC)Reply

Delete CODES

[edit]

I looked at the documentation https://codes.arizona.edu/toolbox/help/. It has a very basic GP regression implementation. The implemenation can optionally act as a DACE wrapper. Some cells in the table seem wrong: solvers:specialized, solvers:exact, and hyperparameters. Joanico (talk) 22:15, 2 March 2025 (UTC)Reply