search

openjournals / joss-reviews

Reviews for the Journal of Open Source Software
Creative Commons Zero v1.0 Universal
721 stars 38 forks source link

[REVIEW]: MultiPrecisionArrays.jl: A Julia package for iterative refinement #6698

Closed editorialbot closed 1 month ago

editorialbot commented 6 months ago

Submitting author: !--author-handle-->@ctkelley<!--end-author-handle-- (Carl Kelley) Repository: https://github.com/ctkelley/MultiPrecisionArrays.jl Branch with paper.md (empty if default branch): Joss_Paper Version: 0.1.3 Editor: !--editor-->@vissarion<!--end-editor-- Reviewers: @vlc1, @rafaelbailo Archive: 10.5281/zenodo.13851500

Status

status

Status badge code:

HTML: <a href="https://joss.theoj.org/papers/acb40a93b83a948fac316f5cf6adb0f3"><img src="https://joss.theoj.org/papers/acb40a93b83a948fac316f5cf6adb0f3/status.svg"></a>
Markdown: [![status](https://joss.theoj.org/papers/acb40a93b83a948fac316f5cf6adb0f3/status.svg)](https://joss.theoj.org/papers/acb40a93b83a948fac316f5cf6adb0f3)

Reviewers and authors:

Please avoid lengthy details of difficulties in the review thread. Instead, please create a new issue in the target repository and link to those issues (especially acceptance-blockers) by leaving comments in the review thread below. (For completists: if the target issue tracker is also on GitHub, linking the review thread in the issue or vice versa will create corresponding breadcrumb trails in the link target.)

Reviewer instructions & questions

@vlc1 & @imciner2, your review will be checklist based. Each of you will have a separate checklist that you should update when carrying out your review. First of all you need to run this command in a separate comment to create the checklist:

@editorialbot generate my checklist

The reviewer guidelines are available here: https://joss.readthedocs.io/en/latest/reviewer_guidelines.html. Any questions/concerns please let @vissarion know.

✨ Please start on your review when you are able, and be sure to complete your review in the next six weeks, at the very latest ✨

Checklists

📝 Checklist for @vlc1

📝 Checklist for @rafaelbailo

editorialbot commented 6 months ago

Hello humans, I'm @editorialbot, a robot that can help you with some common editorial tasks.

For a list of things I can do to help you, just type:

@editorialbot commands

For example, to regenerate the paper pdf after making changes in the paper's md or bib files, type:

@editorialbot generate pdf
editorialbot commented 6 months ago

Software report:

github.com/AlDanial/cloc v 1.90  T=0.03 s (2006.0 files/s, 141935.5 lines/s)
-------------------------------------------------------------------------------
Language                     files          blank        comment           code
-------------------------------------------------------------------------------
Julia                           29            256            772           1306
Markdown                        22            391              0           1231
TeX                              3             44              0            361
YAML                             8             12              9            181
TOML                             3              5              0             31
-------------------------------------------------------------------------------
SUM:                            65            708            781           3110
-------------------------------------------------------------------------------

Commit count by author:

   443  Tim Kelley
   211  C. T. (Tim) Kelley
     5  CompatHelper Julia
editorialbot commented 6 months ago

Paper file info:

📄 Wordcount for paper.md is 1619

✅ The paper includes a Statement of need section

editorialbot commented 6 months ago

License info:

✅ License found: MIT License (Valid open source OSI approved license)

editorialbot commented 6 months ago 
Reference check summary (note 'MISSING' DOIs are suggestions that need verification):

OK DOIs

- 10.1137/23M1549079 is OK
- 10.1137/17M1122918 is OK
- 10.1137/17M1140819 is OK
- 10.1137/1.9780898718027 is OK
- 10.1137/20M1342902 is OK
- 10.1137/1.9781611977271 is OK
- 10.5281/zenodo.4284807 is OK
- 10.48550/arXiv.2307.16051 is OK
- 10.5281/zenodo.7521427 is OK
- 10.48550/arXiv.2311.14616 is OK
- 10.1137/141000671 is OK
- 10.1145/1141885.1141894 is OK
- 10.1137/0907058 is OK

MISSING DOIs

- No DOI given, and none found for title: Improving the accuracy of computed eigenvalues and...
- No DOI given, and none found for title: Progress Report on the Automatic Computing Engine
- No DOI given, and none found for title: Bi-CGSTAB: A fast and smoothly converging variant ...

INVALID DOIs

- None
editorialbot commented 6 months ago

:point_right::page_facing_up: Download article proof :page_facing_up: View article proof on GitHub :page_facing_up: :point_left:

danielskatz commented 5 months ago

👋 @vissarion, @vlc1, @imciner2 - it looks like this review started about a month ago, but nothing has happened since then. Is there anything I can do as track editor to help?

vissarion commented 5 months ago

@danielskatz thanks for the heads up!

@vlc1, @imciner2 are you experiencing any difficulties in your review?

danielskatz commented 5 months ago

👋 @@vlc1 & @imciner2 - I just wanted to check on if you are still able to work on this? (As track editor, I try to check on reviews in the CSISM track where no progress has been recorded in a 2-week period.)

ctkelley commented 4 months ago

Hi, I would like to add a new reference to the paper and a line or two about that new work. If the paper is under review, I can wait to do that until the review is complete. Otherwise I will update the paper.

Please let me know.

Thanks,

— Tim

On Sat, Jun 15, 2024 at 9:06 AM Daniel S. Katz @.***> wrote:

👋 @@vlc1 https://github.com/vlc1 & @imciner2 https://github.com/imciner2 - I just wanted to check on if you are still able to work on this? (As track editor, I try to check on reviews in the CSISM track where no progress has been recorded in a 2-week period.)

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2170088643, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX66I2LANPRPYPYD36ALZHRRCHAVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDCNZQGA4DQNRUGM . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley Department of Mathematics, Box 8205 SAS Hall 2311 Stinson Drive North Carolina State University Raleigh, NC 27695-8205 (919) 515-7163, (919) 515-3798 (FAX) @.*** http://www.math.ncsu.edu/~ctk/

danielskatz commented 4 months ago

@vissarion should answer this as the editor, but in my opinion, it shouldn't be a big deal to do

ctkelley commented 4 months ago

Thanks. I will hold off until I hear from @vissarion

On Wed, Jul 10, 2024 at 11:58 AM Daniel S. Katz @.***> wrote:

@vissarion https://github.com/vissarion should answer this as the editor, but in my opinion, it shouldn't be a big deal to do

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2221220933, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX674UPMZXLNZNHQVYVLZLV73PAVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMRRGIZDAOJTGM . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley Department of Mathematics, Box 8205 SAS Hall 2311 Stinson Drive North Carolina State University Raleigh, NC 27695-8205 (919) 515-7163, (919) 515-3798 (FAX) @.*** http://www.math.ncsu.edu/~ctk/

vissarion commented 4 months ago

@ctkelley sure, please proceed.

vlc1 commented 4 months ago

Review checklist for @vlc1

Conflict of interest

Code of Conduct

General checks

Functionality

Documentation

Software paper

ctkelley commented 4 months ago

Thanks @vissarion . The new version is up there.

vlc1 commented 4 months ago

Dear @ctkelley, thank for this nice addition to Julia's linear algebra ecosystem. I was able to reproduce the example you provide in the JOSS manuscript:

using BenchmarkTools
using MultiPrecisionArrays
using MultiPrecisionArrays.Examples

N = 4096
G = Gmat(N)
A = I - G
x = ones(N)
b = A * x

@btime lu($A)
@btime mplu($A)

AF = lu(A); @btime $AF \ $b
MPF = mplu(A); @btime $MPF \ $b

which resulted in

# lu(A)
  241.440 ms (4 allocations: 128.03 MiB)
# mplu(A)
  120.972 ms (15 allocations: 64.09 MiB)
# AF \ b
  7.387 ms (2 allocations: 32.05 KiB)
# MPF \ b
  236.096 ms (35 allocations: 16.80 KiB)

The factorisation is indeed twice faster, but the solution stage is much slower, the number of allocations is significant (support for the in-place ldiv! would be worthwhile, imho). Is this expected?

Also, while the README.md and JOSS manuscript clearly stipulate that MultiPrecisionArrays.jl targets dense matrices, I couldn't help try it on a sparse matrix,

using LinearAlgebra
using SparseArrays
using BenchmarkTools
using MultiPrecisionArrays

using Base: OneTo

laplacian(n) = spdiagm(-1 => -ones(n-1), 0 => 2ones(n), +1 => -ones(n-1))
laplacian(n, m) = kron(I(m), laplacian(n)) + kron(laplacian(m), I(n))

n, m = 32, 32

A = laplacian(n, m)
x = reshape([i * j / (n * m) for i in OneTo(n), j in OneTo(m)], :)
b = A * x

@btime lu($A)
@btime mplu($A)

AF = lu(A); @btime \($AF, $b)
MPF = mplu(A); @btime \($MPF, $b)

which led to

# lu(A)
  1.492 ms (78 allocations: 1.38 MiB)
# mplu(A)
  7.088 s (25 allocations: 1.40 MiB)
# AF \ b
  22.826 μs (2 allocations: 16.25 KiB)
# MPF \ b
  268.626 μs (30 allocations: 12.83 KiB)

While I wasn't expecting this to perform well out of the box, the performance gap in the factorization stage surprised me. Could you also comment on this? Thank you very much!

ctkelley commented 4 months ago

Thanks @vlc1 for the insightful comments.

I will respond to the two questions separately to avoid one lone confusing reply.

-- Tim

1) Sparse solvers:

The sparse solvers in Julia only support Float64 and iterative refinement is unlikely to happen until they do. So when you try to factor a Float32 sparse matrix, it gets promoted to Float64 before the factorization.

julia> using LinearAlgebra

julia> using SparseArrays

julia> using BenchmarkTools

julia> laplacian(n) = spdiagm(-1 => -ones(n-1), 0 => 2ones(n), +1 => -ones(n-1));

julia> n=100; L=laplacian(n);

julia> LF=lu(L);

julia> LF
SparseArrays.UMFPACK.UmfpackLU{Float64, Int64}
L factor:
100×100 SparseMatrixCSC{Float64, Int64} with 199 stored entries:

julia> LS=Float32.(L);

julia> LSF=lu(LS);

julia> eltype(LSF)
Float64

julia> @btime lu($L);
  27.083 μs (83 allocations: 119.40 KiB)

julia> @btime lu($LS);
  27.333 μs (91 allocations: 125.18 KiB)

So you should not expect any improvement from doing the LU in single. mplu builds the MPArray structure and factors the low precision copy. Since lu will promote that low precision copy, I'd expect the cost of all that to be high. I did not expect it to be as high as it was, but have not tuned the code for sparsity at all.

There is an open issue about this here

https://github.com/ctkelley/MultiPrecisionArrays.jl/issues?q=is%3Aopen+is%3Aissue

Moreover, sparse solvers spend a lot of time in graph computations which are integer computations. That work does not depend on the storage precision. There has been some work IR for sparse problems, but it's been in C++ and Julia does not support what that work did.

ctkelley commented 4 months ago

Hi @vlc1. I really appreciate your finding the problem with the solver time.

In this particular case, the solver takes 5 iteraitons with a call to ldiv! for the low precision factorization and a call to mul! for the matrix-vector product to compute the residual. Each of those is N^2 work, so I'd expect the solve to be (roughly) 10-15 times the cost of ldiv! for the high precision factorization.

I have been able to duplicate your results and there is clearly a performance bug in the solvers. I will fix that and update the repo and let you know when everything is done.

-- Tim

vlc1 commented 4 months ago

Dear @ctkelley, thanks for the detailed answer, that sounds great!

ctkelley commented 4 months ago

Hi @vic1

I found the problem. When I updated the termination criterion I had to compute

optnorm(A, Inf)

which is a very bad idea and costs as much as a few IR iterations. Changing Inf to 1 makes things a lot better and reverting to termination on small relative residuals makes it better still.

So, I have to get this sorted out and add a one or two kwargs to mplu, ... to let the solvers know what do to. I will let you know once this is in there and documented.

— Tim

On Tue, Jul 16, 2024 at 3:14 PM Vincent Le Chenadec < @.***> wrote:

Dear @ctkelley https://github.com/ctkelley, thanks for the detailed answer, that sounds great!

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2231658302, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX63CKZ7SSEEXUJXROETZMVWI3AVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDEMZRGY2TQMZQGI . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley Department of Mathematics, Box 8205 SAS Hall 2311 Stinson Drive North Carolina State University Raleigh, NC 27695-8205 (919) 515-7163, (919) 515-3798 (FAX) @.*** http://www.math.ncsu.edu/~ctk/

ctkelley commented 3 months ago

Hi, @vlc1

I have fixed the performance bug. Thanks again for finding that.

I have also updated the paper a bit with some timing data for the complete solve on lines 121 - 140 of the draft.

I made more substantial changes to the documentation to explain the issues with termination criteria and the relative costs of the factorization and the triangular solves.

If the new version of the paper is acceptable, I will register v0.1.3 with the performance bug fix and the updated documentation included.

-- Tim

danielskatz commented 3 months ago

👋 @vissarion - How is this submission coming? It seems to be a little stalled?

vissarion commented 3 months ago

@vlc1 are you OK with the fix described above by @ctkelley ?

@imciner2 are you still able to work on this review?

@danielskatz I just pinged the reviewers and waiting for responses. Since this is a holiday season in several places in Europe (Greece included, where I am located) I expect a bit of delay in responses.

danielskatz commented 1 month ago

@vissarion - it's been about 5 weeks now, which doesn't seem promising.

vissarion commented 1 month ago

@danielskatz I agree, I also send a personal email to @imciner2 a few days ago without getting any reply, so I will start searching for a new 2nd reviewer.

vissarion commented 1 month ago

Hi @rafaelbailo a few months ago you were interested and available to review this submission, are you still interested? We are searching for a second reviewer here.

rafaelbailo commented 1 month ago

Happy to do it, @vissarion. I'll try to get the review done within a week. 👍

vissarion commented 1 month ago

@editorialbot add @rafaelbailo as reviewer

editorialbot commented 1 month ago

@rafaelbailo added to the reviewers list!

vissarion commented 1 month ago

@rafaelbailo thanks a lot, please follow the instructions in the beginning of this thread https://github.com/openjournals/joss-reviews/issues/6698#issue-2271161297

rafaelbailo commented 1 month ago

Review checklist for @rafaelbailo

Conflict of interest

Code of Conduct

General checks

Functionality

Documentation

Software paper

rafaelbailo commented 1 month ago

@editorialbot generate pdf

editorialbot commented 1 month ago

:point_right::page_facing_up: Download article proof :page_facing_up: View article proof on GitHub :page_facing_up: :point_left:

rafaelbailo commented 1 month ago

Hi @ctkelley (CC: @vissarion), I have just submitted my review comments as an issue in the package's repository. Please let me know when you get the chance to address them, or if you'd like to discuss any of them further.

ctkelley commented 1 month ago

Hi @rafaelbailo Thanks for the thoughtful review. I will respond to your comments no later than the end of next week.

-- Tim

ctkelley commented 1 month ago

Hi @rafaelbailo https://github.com/rafaelbailo . Thanks again for the thoughtful review. Here is a point-by-point reply to your report. I have mostly followed your suggestions. The exception is the "no PR please" statement in the readme. I have explained that in my response

-- Tim

The Paper

  1. Code examples in JOSS papers: I eliminated that section as you suggested.

The Repository

1.

link to documentation in About section: Done 2.

LinearAlgebra.jl: I used your idea about Reexport.jl 3.

Dependencies section: I reworded the part about "install" 4.

Community guidelines:

A) Issue templates: Done

B) I would like to avoid PRs until I get QR and nonlinear solvers finished sometime in the next 6-12 months. These two things are tricky and I do not want to have to deal with PRs at the same time. After that's done, PRs will be welcome.

The Code:

1.

LinearAlgebra.jl alternative: Done 2.

CodeCov has done that to me before. I think it's harmless. 3.

Stray comments: Cleaned up 4.

Formatting: Done 5.

Docs: fixed in item 1 of Repository

On Mon, Sep 23, 2024 at 12:31 PM Rafael Bailo @.***> wrote:

Hi @ctkelley https://github.com/ctkelley (CC: @vissarion https://github.com/vissarion), I have just submitted my review comments as an issue https://github.com/ctkelley/MultiPrecisionArrays.jl/issues/51 in the package's repository. Please let me know when you get the chance to address them, or if you'd like to discuss any of them further.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2368793606, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX632CMLQ67CKVPUJA53ZYA66BAVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNRYG44TGNRQGY . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

rafaelbailo commented 1 month ago

Hi @vissarion, @ctkelley has addressed all of my suggestions and I have ticked every item on my reviewer checklist. As far as I'm concerned, the submission is ready for acceptance. 👍🏻

vissarion commented 1 month ago

@editorialbot remove @imciner2 from reviewers

editorialbot commented 1 month ago

@imciner2 removed from the reviewers list!

vissarion commented 1 month ago

@editorialbot generate pdf

editorialbot commented 1 month ago

:point_right::page_facing_up: Download article proof :page_facing_up: View article proof on GitHub :page_facing_up: :point_left:

vissarion commented 1 month ago

Thanks @rafaelbailo for your quick review!

@ctkelley I have opened an issue regarding references https://github.com/ctkelley/MultiPrecisionArrays.jl/issues/60 and also a PR that fixes some typos https://github.com/ctkelley/MultiPrecisionArrays.jl/pull/59 please let me know if this is OK.

ctkelley commented 1 month ago

That's fine and I have merged the PR

On Wed, Sep 25, 2024 at 5:23 AM Vissarion Fisikopoulos < @.***> wrote:

Thanks @rafaelbailo https://github.com/rafaelbailo for your quick review!

@ctkelley https://github.com/ctkelley I have opened an issue regarding references ctkelley/MultiPrecisionArrays.jl#60 https://github.com/ctkelley/MultiPrecisionArrays.jl/issues/60 and also a PR that fixes some typos ctkelley/MultiPrecisionArrays.jl#59 https://github.com/ctkelley/MultiPrecisionArrays.jl/pull/59 please let me know if this is OK.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2373538278, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX6ZOYMAABZIGX5PXXSLZYJ6ILAVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZTGUZTQMRXHA . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

vissarion commented 1 month ago

I am also OK after parsing the text and all my comments are addressed.

@ctkelley when a submission is ready to be accepted, we ask that the authors issue a new tagged release of the software (if changed), and archive it (see this guide). Please do this and post the version number and archive DOI here. Please make sure that the author names, affiliations as well as the title of the archive (e.g. zenodo) is exactly the same with the submission.

ctkelley commented 1 month ago

Hi, I released and tagged v0.1.3, but am unable to update the Zenodo DOI. When I get to Zenodo it asks me to make a new release. That is what I just did.

What do I do now?

I had a DOI for the repo but I am not clear on how to update that one.

On Fri, Sep 27, 2024 at 5:11 AM Vissarion Fisikopoulos < @.***> wrote:

I am also OK after parsing the text and all my comments are addressed.

@ctkelley https://github.com/ctkelley when a submission is ready to be accepted, we ask that the authors issue a new tagged release of the software (if changed), and archive it (see this guide https://docs.github.com/en/repositories/archiving-a-github-repository/referencing-and-citing-content). Please do this and post the version number and archive DOI here. Please make sure that the author names, affiliations as well as the title of the archive (e.g. zenodo) is exactly the same with the submission.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2378818430, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX66BUY2VUMV2XEXB5X3ZYUOK7AVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZYHAYTQNBTGA . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

ctkelley commented 1 month ago

Could Tagbot have screwed this up? Should I make a 0.1.3.1 release or something like that?

On Fri, Sep 27, 2024 at 11:10 AM C. T. Kelley @.***> wrote:

Hi, I released and tagged v0.1.3, but am unable to update the Zenodo DOI. When I get to Zenodo it asks me to make a new release. That is what I just did.

What do I do now?

I had a DOI for the repo but I am not clear on how to update that one.

On Fri, Sep 27, 2024 at 5:11 AM Vissarion Fisikopoulos < @.***> wrote:

I am also OK after parsing the text and all my comments are addressed.

@ctkelley https://github.com/ctkelley when a submission is ready to be accepted, we ask that the authors issue a new tagged release of the software (if changed), and archive it (see this guide https://docs.github.com/en/repositories/archiving-a-github-repository/referencing-and-citing-content). Please do this and post the version number and archive DOI here. Please make sure that the author names, affiliations as well as the title of the archive (e.g. zenodo) is exactly the same with the submission.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2378818430, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX66BUY2VUMV2XEXB5X3ZYUOK7AVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZYHAYTQNBTGA . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

ctkelley commented 1 month ago

I think I know what's going on. I got a Zenodo DOI for the package a couple years ago and then renamed it. Zendo does not seem to be able to handle that. When I registered v0.1.3 with JuliaHub TagBot automatically made it release 0.1.3 via GitHub actions.

So, Zenodo will let me get a new DOI, but I will have to make a new tag. So, is it ok if I tag v0.1.3.1 but leave the package version a 0.1.3 to be consistent with the Julia registry?

On Fri, Sep 27, 2024 at 11:12 AM C. T. Kelley @.***> wrote:

Could Tagbot have screwed this up? Should I make a 0.1.3.1 release or something like that?

On Fri, Sep 27, 2024 at 11:10 AM C. T. Kelley @.***> wrote:

Hi, I released and tagged v0.1.3, but am unable to update the Zenodo DOI. When I get to Zenodo it asks me to make a new release. That is what I just did.

What do I do now?

I had a DOI for the repo but I am not clear on how to update that one.

On Fri, Sep 27, 2024 at 5:11 AM Vissarion Fisikopoulos < @.***> wrote:

I am also OK after parsing the text and all my comments are addressed.

@ctkelley https://github.com/ctkelley when a submission is ready to be accepted, we ask that the authors issue a new tagged release of the software (if changed), and archive it (see this guide https://docs.github.com/en/repositories/archiving-a-github-repository/referencing-and-citing-content). Please do this and post the version number and archive DOI here. Please make sure that the author names, affiliations as well as the title of the archive (e.g. zenodo) is exactly the same with the submission.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2378818430, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX66BUY2VUMV2XEXB5X3ZYUOK7AVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZYHAYTQNBTGA . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

ctkelley commented 1 month ago

I got it to work with a tag for v0.1.3-joss.

Version = 0.1.3 DOI = 10.5281/zenodo.13851500

Let me know if this is ok.

On Fri, Sep 27, 2024 at 12:13 PM C. T. Kelley @.***> wrote:

I think I know what's going on. I got a Zenodo DOI for the package a couple years ago and then renamed it. Zendo does not seem to be able to handle that. When I registered v0.1.3 with JuliaHub TagBot automatically made it release 0.1.3 via GitHub actions.

So, Zenodo will let me get a new DOI, but I will have to make a new tag. So, is it ok if I tag v0.1.3.1 but leave the package version a 0.1.3 to be consistent with the Julia registry?

On Fri, Sep 27, 2024 at 11:12 AM C. T. Kelley @.***> wrote:

Could Tagbot have screwed this up? Should I make a 0.1.3.1 release or something like that?

On Fri, Sep 27, 2024 at 11:10 AM C. T. Kelley @.***> wrote:

Hi, I released and tagged v0.1.3, but am unable to update the Zenodo DOI. When I get to Zenodo it asks me to make a new release. That is what I just did.

What do I do now?

I had a DOI for the repo but I am not clear on how to update that one.

On Fri, Sep 27, 2024 at 5:11 AM Vissarion Fisikopoulos < @.***> wrote:

I am also OK after parsing the text and all my comments are addressed.

@ctkelley https://github.com/ctkelley when a submission is ready to be accepted, we ask that the authors issue a new tagged release of the software (if changed), and archive it (see this guide https://docs.github.com/en/repositories/archiving-a-github-repository/referencing-and-citing-content). Please do this and post the version number and archive DOI here. Please make sure that the author names, affiliations as well as the title of the archive (e.g. zenodo) is exactly the same with the submission.

— Reply to this email directly, view it on GitHub https://github.com/openjournals/joss-reviews/issues/6698#issuecomment-2378818430, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACOEX66BUY2VUMV2XEXB5X3ZYUOK7AVCNFSM6AAAAABHABDDT6VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDGNZYHAYTQNBTGA . You are receiving this because you were mentioned.Message ID: @.***>

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

-- C. T. Kelley 2801 Glenwood Gardens Ln Unit 302 Raleigh, NC 27608 @.***

vissarion commented 1 month ago

@editorialbot set 10.5281/zenodo.13851500 as archive

editorialbot commented 1 month ago

Done! archive is now 10.5281/zenodo.13851500