Update to output neuro/conmat datatype.

[soichih]

Is it possible to update this App to output the neuro/conmat datatype like "fMRI to Connectivity Matrices" App? (https://brainlife.io/app/5c720cf63e2f2c0030a23486)

Here is the datatype description

https://brainlife.io/datatypes/5d34d9f744947d8aea0e0d2f

cm.csv and label.json should look like this.

[bcmcpher]

The .csv is easy enough to set up - it writes that now.

But how am I supposed to handle multiple / optional outputs? Currently this can make up to 2 different edge weights (density, average property), but in theory it can scale to many more than this (count, length, density corrected for length, multiple tensor/noddi values). Can these just be tagged appropriately? Can we even work with optional outputs yet?

And there's nothing I can do about label.json in this app. This has to come with the parcellation that is passed. There is no way for me to figure out what the labels are supposed to be when I only get parc.nii.gz.

[soichih]

I believe other outputs can be added as optional output for neuro/conmat (right now there is only cm.csv that's listed). If there are standard measures that should be required across other connectivity Apps, we can make them required as well. Right now this datatype is used by fMRI to connectivity matrices app. I will ping @faskowit about this.

You can convert the key.txt that's part of parcellation/volume input to label.json. Like this

https://github.com/faskowit/app-fmri-2-mat/blob/0.0.4/generate_cm_datatype.py#L12

Please note.. as part of our datatype discussion about parcellation/volume, I am planning to propose parcellation/volume datatype to use label.json format instead of key.txt. When this happens, you should just have to copy the label.json into neuro/conmat datatype instead of parsing key.txt.. I will push for that soon.