-
Notifications
You must be signed in to change notification settings - Fork 7
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[Re] Measures for investigating the contextual modulation of information transmission. #29
Comments
I think that is everything required to start the process. This authors of this paper used classical information theory. The nascent field of Partial Information Decomposition (PID) enables breaking three-way (and some formulations with more terms) information theory into more precise parts. Since this is a journal for replicating results of old studies, I have limited the scope to that and not discussed any of this in the submission or included any PIDs here, but that can change if desired. |
@sepehrmn Sorry for the delay, the Ten Years Reproducibility Challenge has made us quite busy these past days. @sepehrmn For figure 2, yu might want to put sub-figures side by side and try to fix the aspect ratio. Currently, they're a bit difficult to read. @gdetor @oliviaguest @benoit-girard @eroesch Can any of you take care of this regular submission in computational neuroscience ? |
Happy to take this on. 😄 |
@oliviaguest If you don't mind, I'd be happy to act as a reviewer for this submission. Would be my first one then. 🙂 |
@oliviaguest Thank you @schmidDan Thank you, I guess you can start your review but @oliviaguest will confirm it. |
@schmidDan given you have only a couple repos and they are both forks, can you link me to your institutional profile and/or website, please? It will help me decide on who else to invite and understand (of course) your skills. Thanks. |
@oliviaguest thanks for the fast reply! Since I'm starting out with my PhD studies at the moment, my institutional website is still to become available. My affiliation, though, is with the Institute for Neural Information Processing, Ulm University, Germany. The closest thing I can link for reference would be my ORCID. Regarding public repos you're right. So far repos showcasing my work aren't available to the public. Something I hopefully can change throughout the year. |
@schmidDan you are super picked — thank you so much for the info. 😊 |
@oliviaguest Thanks a lot! Then I'm more than happy to aid as a reviewer. 🙂 |
Excellent! I am very happy to see that you have taken charge of the reviewing process Olivia and Daniel. I'll see what I can do to improve figure 2 as Nicolas suggested by tomorrow. |
@sepehrmn I'm the editor. I'll find you a 2nd reviewer soon hopefully. Any suggestions welcome, of course! |
Thanks for providing some preliminary feedback. Figure 2 has now been adjusted accordingly. |
Perhaps due to the current crisis I am struggling a bit to find a 2nd reviewer as quickly as I would like. @sepehrmn if you have any ideas, they are welcomed. I will keep looking of course. |
👋 Hi @pwollstadt @Abzinger @pmediano @finnconor — would any one of you be interested in reviewing this? 😊 |
@oliviaguest I’m interested … :) |
@oliviaguest I'm interested. I just want to note that I'm currently working with Sepher in the same lab. But I haven't been part of the work he is submitting (no involvement in the discussion, the techniques used, the implementation of the code, the results obtained, or the written manuscript). As a reference, I suggest looking at my ORCID |
OK, fantastic. I suggest you both (@reubsjw @Abzinger) act as 2nd and 3rd reviewers. The rest of the people tagged above can have a chance (if they so wish) to review in future. Great! If all three @reubsjw @Abzinger @schmidDan can give me a rough ETA for your reviews that would be very helpful. Please also, if not already, take a look at: http://rescience.github.io/edit/ |
I’ll aim for tomorrow or Saturday if that’s ok? |
The author replicates the target article's findings by implementing the same mathematical models in a modern computational environment. The target article describes and demonstrates, in information-theoretical terms and using computational simulation, additive and modulatory relationships (seperately and jointly) of receptive and contextual inputs to a hypothetical neuron, with that neuron's output, in terms of the information transmitted. My only major concern is the commenting in the code, which includes, e.g. "to-do's" (presumably as an aide memoir to the author). The commenting should also, where possible, reference the numbered equations in the article, but I would also encourage them to provide short explanations for each block/loop etc of the code (which they have already done nicely in many places), to help readers orient themselves. It would also be nice if the statement that the original paper is "...quantitatively replicated..." was qualified a little more. I note that Figure 1, for example, appears to be noticably different (albeit the qualitative shapes of the lines are the same, and the layout of the figure in Smyth et al is harder to read than in the present replication). Also, speaking of figures being easier or harder to read, I'd like to recommend that for figure 2, the author use the same "perspective" (as it were) on the 3D plots as is used in the Smyth article, since that layout is much more readable in my opinion. |
@oliviaguest ETA would be by the end of today or tomorrow for my review. |
@oliviaguest At the latest, I will have my review done by the end of the day |
Gosh, you're all super fast. Thank you. 😱 😍 |
@sepehrmn I had a fun time reading and reviewing your code and manuscript. Nice job! Please find my review below. My review is based on commit 2c117cf. Review Summary
Remarks w.r.t. the manuscript
Remarks w.r.t. the implementation
(edit: added link to commit I based my review on) |
@sepehrmn It was enjoyable going through the manuscript and the implementation. Below is my review which is hopefully useful to improve the work. Overview
CommentsHere are some suggestions to improve the manuscript and the code. Manuscript
ImplementationAs a general comment, clean the code of any internal comments like #ToDo or something code that you commented out. Below are two main issues that I think need improving.
|
I'm not sure about the X11 forxwarding part. Best would be to clone the repo into another place and check if they're the same (make lasy commit). |
@oliviaguest The error message But... you can't just run What I don't understand is why |
Yes, due to the "bug" (?) I described in #29 (comment). I added a print statement. I didn't understand how to fix it and ran it with the |
@rougier I don't see a difference [using software tools not my eyes] between my local version and a fresh clone. Sorry. What can I do to fix this? I feel like this is the same bug I had last time (although it was not from my MBP, but from a Linux desktop). |
OK, I did this manually. I think it's OK! |
Excellent. Many thanks everyone! |
@sepehrmn sorry about the teething problems — I seem to always find a way to screw up right before the end... 😆 BUT! Importantly your paper is published and I think it should appear on the website too soon. So... 🥳 congratulations! 👏 |
Did you add the bibtex entry on the website bibfile? |
No, what is needed to be done? Just copy-paste this one? https://github.com/ReScience/articles/blob/master/10.5281_zenodo.3885793/article.bib |
Are the keywords OK? Hmm... See: ReScience/rescience.github.io@2f3b5b6 |
Perfect. And yes, you only need to append it to the website bibfile (I'll regenerate the website) |
Great! Excited to see @sepehrmn's paper on the https://rescience.github.io site. 👏 |
Excellent! Much appreciated @oliviaguest ! I guess the issue can be closed now! |
Let's make sure everything is OK first. |
Very well. I still can't see it on https://rescience.github.io/read/ . |
@rougier now sure what's up? Have you compiled and it's my manual editing of files that's screwy? |
@sepehrmn feel free to unsubscribe BTW. We will deal with this as it's a journal issue — congrats again and thank you for being so patient! 🥳 |
@oliviaguest I'm a bit lost now. What do I need to check exactly ? |
@oliviaguest is it supposed to be published within the "Issue 2 (NeurIPS 2019 Reproducibility Challenge)" though? |
@schmidDan I am confused! I think I did something wrong again! @rougier? ReScience/ReScience#48 (comment) |
Probably the issue number is wrong. I think it should be 3 and paper number should be #2. You can correct it directly on Zenodo and you'll need to update the published.bib on the rescience website (or I can do it, just tell me). |
@rougier looking on Zenodo, maybe I'm confused... but it looks right? I haven't changed anything. |
Ok, it's only wrong in the bibtex for the website then. I'll correct it. |
Oh! Thank you, @rougier. 🌷 |
Done ! |
Rejoice! Thanks! 🥳 |
Original article:
Smyth, D., Phillips, W. A. and Kay, J.(1996) 'Measures for investigating the contextual modulation of information transmission', Network: Computation in Neural Systems, 7:2,307 — 316
PDF URL:
https://github.com/sepehrmn/mahmoudian-2020-rescience/blob/master/article/Reproduction_of_Smyth_et_al__1996.pdf
Metadata URL:
https://github.com/sepehrmn/mahmoudian-2020-rescience/blob/master/article/metadata.yaml
Code URL:
https://github.com/sepehrmn/mahmoudian-2020-rescience/tree/master/code
Scientific domain: Computational Neuroscience
Programming language: Python
Suggested editor:
The text was updated successfully, but these errors were encountered: