Update "Further Resources" links

cran-task-views / Econometrics

CRAN Task View: Econometrics

https://CRAN.R-project.org/view=Econometrics

30 stars 4 forks source link

Update "Further Resources" links #5

Closed grantmcdermott closed 2 years ago

grantmcdermott commented 2 years ago

@zeileis @tappek as a general question, do we want to go through PRs for edits to the page? Or, are we fine to add commits directly? I'm fine to adopt the former approach if it's easier, though I do like the PR + review approach for completeness.

A few small comments on these additions:

I changed the "Links" header to "Further resources" since that feels more intuitive to me. Happy to change back if there's some requirement that needs to be met.
While I tried to be judicious, this list is growing quite long. It might be be worth thinking about some additional cutoff criteria to keep this "curated" list exactly that. E.g. I included some books that would also find a natural home in the proposed "Causal Inference" TaskView. We may also want to think about a publish date threshold. I must confess that I have not read the JSS article, but if it's survey of the R econometric landscape circa 2008 then it may be a bit long in the tooth.

zeileis commented 2 years ago

I'm still new to all the GitHub features and hence typically prefer direct commits over pull requests for simple changes like this. But I'll try to learn. Are comments like this what I should do as a reviewer - or should I do something else instead or in addition?
The links section should be formatted exactly as ### Links, it is handled specially in ctv2html().
All the links you added look good to me, I don't think that the list is too long.
As for the old publications: I'm biased, of course, but I think that the papers from the 2008 special JSS volume are still worth reading (on plm, forecast, np, vars, count regression, etc.). But maybe we can move the bullet point to a less prominent position. Similarly, our AER book is still not obsolete, I think, even if a 2nd edition is very much overdue...

grantmcdermott commented 2 years ago

Quick responses:

One nice thing about GitHub's reviewer workflow is that you can comment on specific sections of code, or make edits yourself. It's probably overkill for minor tweaks and edits. But it could be nice for larger section rewrites. I'm happy to hop on a Zoom if you guys ever want to go through some live examples together.
Cool, I just reverted to the "Links" header.
Good to know. I should read the article (ofc). One suggestion that might address your position prominence comment is to add a "Journal" field. We could then include JSS, JOSS, and the R Journal, and specifically highlight the special issue of the former. OTOH, now I'm making the list even longer ;-)

I think I'll go ahead and merge this PR then, that last comment notwithstanding.

zeileis commented 2 years ago

Merge: Thanks, looks good.
Reviewer workflow: I read through this a little bit but apparently not in enough detail. I didn't see how I make additional modifications/commits in the code that is reviewed. Do I make a regular commit+push and then mention the PR ID in the commit message or something like that?
I agree that listing every individual paper related to econometric software in R would be overkill. I listed the special volume as it contains seven such papers plus intro paper. So I thought that it would be justifiable to keep it in the list.