JOSS paper

[naik-aakash]

• Finalize associated text
• Cite other works using this package

[naik-aakash]

Hi @ajjackson , @QuantumChemist and @kaueltzen , It would be great if you can have a look at the updated documentation and the draft prepared for the submission to JOSS.

@ajjackson: We know you are on parental leave, so kindly do it at your convenience. No rush.

I am looking forward to your comments, suggestions and corrections if you spot any erros 😄

[kaueltzen]

Hi, I sent my suggestions for the paper draft directly to @naik-aakash .

Regarding the documentation: it's very detailed and well structured, just one small thing: in https://jageo.github.io/LobsterPy/fundamentals/index.html the first COHP plot is referenced as COHP data for a site and later on it says Similarly, the antibonding percentage is evaluated for the site from the COHP data. etc.
Although the data is stored site-wise, I would clarify here that the plot contains "COHP data for a site and a specific neighbor / bonding partner", i.e. something "bond-wise", not "site-wise". Otherwise it could confuse new users with no prior experience in bonding analysis.

[naik-aakash]

Thanks @kaueltzen for the suggestion . I will make the changes 😃

[QuantumChemist]

Hey! I agree with Katharina and I also found one typo on Getting started (LOBTSER), as well as Fundamental Aspects, I'd rather say "summed COHPs" instead of "summed cohps". My paper suggestions go directly to @naik-aakash as well :)

[naik-aakash]

Hey! I agree with Katharina and I also found one typo on Getting started (LOBTSER), as well as Fundamental Aspects, I'd rather say "summed COHPs" instead of "summed cohps". My paper suggestions go directly to @naik-aakash as well :)

Hey thanks for spotting the mistake as usual. 😄 . Will fix it

[naik-aakash]

Hi @kaueltzen and @QuantumChemist, I re-read the fundamentals section. But in steps 1 and 2, it is made clear the COHP data is explicitly for the bonds at the site. I think it should be clear to the users from this that COHP data is for the bonds at the site.

But if you prefer any specific way to phrase it in the steps, please share exactly what line should be changed or simply raise a PR.

[naik-aakash]

Revised draft

[naik-aakash]

Hi, I sent my suggestions for the paper draft directly to @naik-aakash .

Regarding the documentation: it's very detailed and well structured, just one small thing: in https://jageo.github.io/LobsterPy/fundamentals/index.html the first COHP plot is referenced as COHP data for a site and later on it says Similarly, the antibonding percentage is evaluated for the site from the COHP data. etc. Although the data is stored site-wise, I would clarify here that the plot contains "COHP data for a site and a specific neighbor / bonding partner", i.e. something "bond-wise", not "site-wise". Otherwise it could confuse new users with no prior experience in bonding analysis.

Thanks ! Has been addressed now as part of #208

[naik-aakash]

Hi @kaueltzen and @QuantumChemist, I re-read the fundamentals section. But in steps 1 and 2, it is made clear the COHP data is explicitly for the bonds at the site. I think it should be clear to the users from this that COHP data is for the bonds at the site.

But if you prefer any specific way to phrase it in the steps, please share exactly what line should be changed or simply raise a PR.

[JaGeo]

I made some comments. I hope it helps!

[JaGeo]

Hi @ajjackson ,

We are thinking about submitting the JOSS in one of the next weeks. Would you have time to agree on the paper draft and submission in the next two weeks?

JG

[JaGeo]

@ajjackson , thank you so much!

[naik-aakash]

Hi @JaGeo , @ajjackson , @QuantumChemist and @kaueltzen , I have just added the DOI in the references which were missing earlier and very small change in text where I cite atomate2, instead of URL.

Here is the updated draft PDF

Todo

• Before submission to JOSS, need to remove the paper compiling workflow from this PR

[ajjackson]

Hi everyone, thanks for moving on with this and sorry I haven't been around much lately.

This looks pretty good already 😄 If time allows I might have a go at twiddling some words around, would the best way to do that be as PR to Aakash's branch?

I think the Statement of Need has the right emphasis: the package supports high-throughput studies (with citation of this application) and is also intended to also support ML. Both of those things require transformation of data into a consistent, machine-friendly form and that's what this package is for. In some ways for JOSS review this is the most important part of the manuscript to get right.

One area worth a quick discussion:
I don't think the current manuscript makes it entirely clear what the scope of this package is within the ecosystem of other packages it works with.

We describe it as a package that "provides convenient tools to systematically analyze, describe and visualize such LOBSTER computations results." It's unclear from this quite what we mean by "analyze" -- particularly, whether:

• lobsterpy implements the "LobsterEnv" spatial assignment (that code is actually in Pymatgen)
• there are some arbitrary/expert choices that lobsterpy makes in its data processing (I think it is just the threshold to select significant bonds?)

A more complete version of my affiliation is: Scientific Computing Department, Science & Technology Facilities Council, Rutherford Appleton Laboratory, Didcot, 0X11 0QX, UK (I think that's the same as the ChemPlusChem paper). If it's important to save space the laboratory and postcode can be dropped but it's important to keep the department name.

[naik-aakash]

Hi everyone, thanks for moving on with this and sorry I haven't been around much lately.

This looks pretty good already 😄 If time allows I might have a go at twiddling some words around, would the best way to do that be as PR to Aakash's branch?

I think the Statement of Need has the right emphasis: the package supports high-throughput studies (with citation of this application) and is also intended to also support ML. Both of those things require transformation of data into a consistent, machine-friendly form and that's what this package is for. In some ways for JOSS review this is the most important part of the manuscript to get right.

One area worth a quick discussion: I don't think the current manuscript makes it entirely clear what the scope of this package is within the ecosystem of other packages it works with.

We describe it as a package that "provides convenient tools to systematically analyze, describe and visualize such LOBSTER computations results." It's unclear from this quite what we mean by "analyze" -- particularly, whether:

• lobsterpy implements the "LobsterEnv" spatial assignment (that code is actually in Pymatgen)
• there are some arbitrary/expert choices that lobsterpy makes in its data processing (I think it is just the threshold to select significant bonds?)

A more complete version of my affiliation is: Scientific Computing Department, Science & Technology Facilities Council, Rutherford Appleton Laboratory, Didcot, 0X11 0QX, UK (I think that's the same as the ChemPlusChem paper). If it's important to save space the laboratory and postcode can be dropped but it's important to keep the department name.

Hi @ajjackson , thanks for checking it out already. I have sent an invite to collaborate in my repo fork, so you can directly add changes to this PR.

Regarding your question about "analyze", Agreed that LobsterPy uses LobsterEnv from pymatgen, but through this package we also provide the outputs that are way easier to understand in the form of text description and jsons besides the choices set for the thresholds to select significant bonds. Also means to getting plots automatically which help in interpreting the results to the user.

Also, now we added the option of get_lobster_calc_quality_summary, which in principle reads outputs of calculations to provide a summary of how good the LOBSTER calculation is, which I also considered as an "analyze" aspect.

No idea at the moment how else to phrase it in a better way. If you have any suggestions for this, please let me know, would be happy to make the changes or have a go at twiddling the phrase if you get some time in the next days.

Either way works fine 😄

But anyway, thanks again!!

[JaGeo]

Just to add here: it might be still worth adding that it works very well together with the Materials Project ecosystem :). I think we should mention this as well! Thanks, @ajjackson .

You can, of course, also make a direct review of the paper, if this is easier.

[naik-aakash]

Hi all , I have updated the text to better explain where LobsterPy as a package fits in. Hope this helps to clarify the points raised by @ajjackson

Here is the Updated PDF

[QuantumChemist]

Hi all , I have updated the text to better explain where LobsterPy as a package fits in. Hope this helps to clarify the points raised by @ajjackson

Here is the Updated PDF

I've sent my suggestions directly to Aakash :)

[kaueltzen]

Hi all , I have updated the text to better explain where LobsterPy as a package fits in. Hope this helps to clarify the points raised by @ajjackson
Here is the Updated PDF

I've sent my suggestions directly to Aakash :)

Same

[naik-aakash]

Updated PDF

[JaGeo]

@ajjackson Aakash has now updated the draft. Let us know what you think.

If you have ideas for reviewers, we would be happy to receive them via mail, for example.

Thank you!

[ajjackson]

I'm happy for this to be submitted, I think those points are made clear now 😄

[JaGeo]

Awesome! Thank you so much, @ajjackson !

We will think about reviewers, release a new version and then submit! 😃