Closed imkemayer closed 2 years ago
Hi! Thanks for inviting me to this. I hope commenting here is the right way to interact.
Many of the bullet points refer to areas I don't know much about, but some are for ones I do! Some notes:
lm()
, or heck even mean()
or aggregate()
or a million other functions. If you mean that there aren't basic R functions that directly implement standard CI designs with all the bells and whistles, that's more accurate.lm()
, so I'm not sure this is a useful entry there. Unless I'm missing a specific application they have.rdrobust
is the main one you'd use. Also suggest the addition of RDHonest for inference under poorly-behaving running variablesHi everyone, especially thanks to @imkemayer for starting the discussion here and @NickCH-K for providing detailed feedback!
My first impression is that this looks very promising. The number of packages is ok but if there were something to sharpen the focus, it might help and facilitate maintenance in the future.
Nathalie @tuxette said that she will try to have a closer look soon and provide some more feedback. But the discussion and decision may take some more weeks as we are currently still very busy with the transition of the old task views to the new infrastructure (as mentioned in the README).
Hi @imkemayer. Can you post the source file somewhere on github so that I can fork the repo to propose a few typo corrections? In addition, I am not sure (since I do not have the source) but I think that either you are using the old (ctv) format or you have a problem with your links? Deeper feedback (not only restricted to format) is coming soon.
Thank you @NickCH-K for this great feedback and the suggestions that we will use to update the current proposal!
And thank you @zeileis for your suggestion to sharpen the focus of the task view. We will discuss internally and await also @tuxette's comments to see how we could improve the proposal to become a well maintainable task view.
To answer @tuxette's question about the format, indeed we are still using the old format because we have started working on this last year before knowing the new format. But we will migrate everything to the new format very soon.
My feedback:
Ref to pull request: https://github.com/imkemayer/causal-inference-taskview/pull/1
Thanks for the productive discussion so far and especially to Nathalie @tuxette for having a closer look and writing up her feedback. Some additions:
Imke @imkemayer, now that we have officially relaunched the task views workflow on CRAN, we can make progress on new task view proposals as well. Hence I wondered what the status of this proposal is.
@zeileis We have a meeting planned in two weeks to improve it again and I think we will have another contributor/maintainer as well. Keep you inform asap.
OK, thanks!
I am a bit late to the party, but just a small comment: The current state already looks very comprehensive and well-structured to me. Shouldn't there be a reference to the graphical models task view? (pcalg package and graphical models are briefly mentioned).
Thanks @davidjohannesmeyer for this comment, indeed that reference to the graphical models ctv should be added to complete the part on the causal structure learning. This has been done now and can be consulted in the updated preview of the task view proposal: https://misscausal.gitlabpages.inria.fr/misscausal.gitlab.io/files/ctv/CausalInference.html
Hi @imkemayer ! Do you consider the last update of the TV as a new proposal? There has not been much activity in the current issue during last month but I don't think that all remarks from @zeileis and myself have been addressed. It does not mean that they all need to be taken into account but a point-by-point answer here would be useful to understand your point and make this TV move forward. Thanks for your work!
Hi @tuxette, thank you for the reminder. Indeed we haven't addressed all comments and suggestions yet but we will do it this week together with @julierennes and come back to you with a point-by-point answer.
Thank you all for your suggestions and remarks. As suggested by @tuxette, we provide a point-by-point reply to your remarks:
@NickCH-K (adressed by commit #a48e507)
@tuxette and @zeileis (addressed by commits #9231acf and #9231acf)
Thanks for the corrections and the point-by-point answer. This clarifies things for me. As a minor remark, I might have missed something but wouldn't it be best to directly include @NickCH-K as one of the co-maintainers?
Indeed @tuxette, we have not included him yet but finally did it in the most recent commit #8bec5d6 Thanks for pointing this out to us!
I am under the impression that we could move forward to the publication of the TV: what do you think, @zeileis ?
Given they have reached out from someone in political science, I think it would be nice to wait for their feedback.
But after that we can quickly move forward. Formally, we need a third endorsement from one of the CRAN Task View Editors. A thumbs up reaction would be enough, @rsbivand @eddelbuettel @rociojoo @davidjohannesmeyer
What is the current URL? For https://github.com/imkemayer/causal-inference-taskview/ I get a 404.
Does this work for you: https://www.imkemayer.com/files/causalinference ? Sorry for not checking that before...
One small comment and question about the instrumental variables section in the task view:
ivmodel
and ivpack
listed explicitly rather than ivreg
? I'm biased here obviously...possibly I'm overlooking important advantages of these packages.view("Econometrics", "Instrumental variables")
now. This requires the very latest development version of ctv
from R-Forge, though (committed a couple of minutes ago).Thank you @zeileis for pointing out this missing package. Sorry for the oversight, we missed it when searching by key words for potential packages to review. I fixed the search script and now it finds your package.
And thanks for offering the new option to directly refer to a specific section of another task view. We now use it for IV and RDD.
Thanks for all your work Imke @imkemayer & Co! I've now completed the transfer of the repository to the cran-task-views org and did some touch-ups, see: https://github.com/cran-task-views/CausalInference In particular:
So I think we are good to go for a CRAN release. Imke, maybe you can have a quick last look whether everything is in order? Then I'll release on CRAN.
Thank you Achim @zeileis for this quick and complete touch-up of the transferred repository! I've had a look at the changes and they're all fine for me. Also I merged the data and R directories as you suggested.
Fantastic, we're live: 🎉
https://CRAN.R-project.org/view=CausalInference
And advertised on Twitter: https://twitter.com/AchimZeileis/status/1540329721623347200
Thanks everyone for your efforts!
Scope
This CRAN Task View proposal contains a list of packages useful for causality related analyses, including causal discovery and mainly causal inference. On top of the list of reviewed packages, we plan to include a list of selected books and articles that make use of some of these packages and that are helpful to get a better understanding of the topic of causality on concrete examples implemented in R code. Causal analyses can be performed in many different ways, depending on the context, the data and the assumptions made by the analyst. With this task view we aim at assisting the user in the choice of adequate implemented methods for his/her problem and data to analyze.
Since some of this proposed CTV maintainers are also maintainers of the already existing MissingData CTV, we followed a similar approach to select packages into the task view. To be considered for inclusion, the package must be useful for running causal analyses, be it causal discovery, causal inference or causal prediction, either by providing analysis tools or relevant datasets. Furthermore the package should satisfy the following technical criteria: the date of first release of the package should be more than one year old and the date of last update or the frequency of updates should reflect a regular maintenance of the package.
Packages
Please see https://www.imkemayer.com/files/causalinference for a comprehensive list and preview of the Task View. It currently consists of 150 packages (including 9 core packages).
Overlap
There might be some overlap between packages in the ClinicalTrials and ExperimentalDesign CTVs, but to minimize this overlap, our CTV proposal focuses on issues that arise mainly in observational (or quasi-experimental) data analysis and less on experimental design. Since causal inference is an important topic in econometrics, there is also some overlap with the Econometrics CTV but we aim at targeting an audience from a wider range of domains, for example with problems from social and health care applications. Missing data handling and causal inference share some similar methods and certain packages can be used in the two contexts but with different goals. Therefore overlap between this present proposal and the MissingData CTV is possible but does not imply redundancy.
Maintainers
The primary maintainer is Imke Mayer (@imkemayer). Current co-maintainers are Pan Zhao (@panzhaooo), Nick Huntington-Klein (@NickCH-K) and Julie Josse (@julierennes).
Collaborators of these maintainers have already agreed to assist in the maintenance of the task view to ensure constant updates and corrections where necessary.