Contributing Fast-SL ALgorithm to COBRA Toolbox

24 views
Skip to first unread message

Somya Verma

unread,
Feb 1, 2023, 12:36:07 AMFeb 1
to COBRA Toolbox
Hi all,

I have been trying to contribute the Fast-SL algorithm to the COBRA Toolbox repository.

I have followed the following steps to make the pull request -

1. Created a Fork Repository.
2. Carried out the tests mentioned to verify the code.
3. Made the pull request.

Despite completing the instructions, the pull request is still denied. I'd appreciate it if you could help me with this.

Thanks and Regards,
Somya

Thomas Pfau

unread,
Feb 1, 2023, 1:25:03 AMFeb 1
to cobra-...@googlegroups.com

Hi,

A couple of things I notice with the PR:
- there is an "untitled.m" file in the PR, should npot be there
- pathes should not have spaces in the toolbox, since this causes a lot of issues in general try to use lower case pathes
- the models that are supplied with the PR should probably be added to the cobra models repository and not added to the main cobra repository
- Is there any reason, why the parallel version cannot be used with other solvers? In general except for not using solveCobraLP but explicitly using a cplex version, I don't see a reason for this to only be available for cplex, as it simply uses parfor
- some more in line commenting and/or some clearer variable names (instead of maybe synthetic_lethals instead of Jsl, synthetic_double_lethals/triple_lethals instead of Jdl/Jtl etc) would help readability of the code a lot.
- The examples could probably be moved to tutorials, along with the information in the README.md.
- The comments should be formatted according to the style guide in order to allow automatic documentation generation for the functionalities
- there are at least 2 fastFL.m files, which can lead to the wrong one being used.
- the Rxns folder conains several of the models that are also in the Models folder. as mentioned above all of those should (if they are not already there) be moved to the models submodule (if they are even necessary for the code.

Best

Thomas

--

---
You received this message because you are subscribed to the Google Groups "COBRA Toolbox" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cobra-toolbo...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cobra-toolbox/7ee3ed06-1e4b-42fb-a29e-3d26ca927995n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages