Create a guideline for the code!

[morteza-mahdiani]

The initial version of the code is pushed to the repo. Update readme to explain how we should set parameters for using the code!

[morteza-mahdiani]

I tried to write sth that can be helpful :)
Let me know if further information is needed! @jwalbrin

Soon I need to add a lot of facilities and other methods to the Connectivity class!

[jwalbrin]

Great. I had a look at the script earlier, but lost a lot of time setting up git. But hopefully will be able to make some edits tomorrow!

…

On Thu, 23 Jun 2022 at 15:19, Morteza ***@***.***> wrote: I tried to write sth that can be helpful :) Let me know if further information is needed! @jwalbrin <https://github.com/jwalbrin> Soon I need to add a lot of facilities and other methods to the Connectivity class! — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXAPNWVCE3DK7DKHPDNQRYLVQRXAHANCNFSM5ZSIUFVQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[morteza-mahdiani]

Thanks, Jon. Let me know if it is needed for a meeting to fix the issues with Github! Actually, the code needs to be modified for sure but just an initial version to check whether we are in the correct way or not! @jwalbrin

[jwalbrin]

Hey Morteza, Nice start with the connectivity script, thanks for sharing it. I tried running the following: c_instance = Connectivity('/mnt/EXT4_DRIVE/MRIData/ETDR/ResidualTimeCourse_THBFP_FIR/',... '/mnt/EXT4_DRIVE/MRIData/ETDR/ROIs/SubjReg_SearchSpaces_GM_ASMasked/',... '/mnt/EXT4_DRIVE/MRIData/ETDR/Analysis/DirectionalConnectivity/Morteza_Class_Test',... 8,10,{'rOFA','rFFA','rSTSF'},... '/home/jon/matlab/Toolboxes/MVGC1-master'); c_instance.preprocess(1100); gc_instance = GCParameters(1,1100,'OLS', 'LWR', 'AIC', 20, 'F', 0.05, 'FDR', 0); c_instance.GCM(gc_instance, 20, '/mnt/EXT4_DRIVE/MRIData/ETDR/Analysis/DirectionalConnectivity/Morteza_Class_Test/'); It mostly runs ok, but I encountered these issues: - path issue with the MVGC toolbox ( I manually changed this and it worked) - the output preprocessed data goes to the penultimate directory in the output path (rather than the final directory) - GCM returns an empty matrix (after preprocessing has occurred; no errors shown, but is of the correct size given nSubj and nReg) A few thoughts on structure: - I think we probably don't want to store raw connectivity data (e.g. as .mat files that are later called) but instead to load this up GC calculation only - And of course we will sometimes want region-wise time-courses, other times voxel wise courses so perhaps 2 alternative functions to implement these Let me know once these issues are corrected and I'll take another look (perhaps it is easier to make commits etc. once we've polished things a little further). Also, your name should go first in the readme as this is primarily your work! Thanks, Jon

…

On Thu, 23 Jun 2022 at 15:37, Morteza ***@***.***> wrote: Thanks, Jon. Let me know if it is needed for a meeting to fix the issues with Github! Actually, the code needs to be modified for sure but just an initial version to check whether we are in the correct way or not! — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AXAPNWRT37K526GWTNQBPUDVQRZEHANCNFSM5ZSIUFVQ> . You are receiving this because you were mentioned.Message ID: ***@***.***>

[morteza-mahdiani]

Hey Jon, Many thanks for the feedback again! I tried to fix the reported bugs if I understood them correctly. I also modified the readme.

About suggestions:

• I am not sure about not storing the preprocessed data as we are adding more functions in the future!
• You are right! I will add other functions or upgrade the existing ones for the voxel-wise analysis.

[morteza-mahdiani]

The option to save or not is available with the saving flag parameter and region-wise/voxel-wise are available now!