Create a comprehensive set of example data

[rob-luke]

Both raw and processed data

(as discussed in sfnirs conference)

[fangq]

hi @rob-luke and @dboas, with the changes I just made in NeuroJSON/jsnirfy@e5e3fb5 and NeuroJSON/easyh5@2090320, all snirf files in the snirf-samples now can pass the validator - the only exception is the minimum_example.snirf but I believe it is due to a bug in the validator (which I filed here: BUNPC/pysnirf2#49).

I think all of you should have write permission to snirf-samples. feel free to add additional files. I suggest keeping the data size relatively small, but keep the data/file type diverse. it would be great for software testing purposes.

Currently, a JSON file (.jsnirf) is serving as the "source-code" for each sample data, and use a matlab script to convert it to .snirf. I feel this way, people can use the json file to understand the data structure, but once converted to .snirf, it is no longer human readable. I am happy to create these .jsnirf files if you upload new .snirf files.

[OliverF2022]

Could you provide an example using nirs[i]/measurementList[i]/dataTypeLabel? I couldn't find any example files that used this label. I'm not quite sure how this label helps confirm the information in dataTimeSeries. Thanks. @fangq

[fangq]

measurementList(i).dataTypeLabel is a string specifying what type of (processed) data stored in the dataTimeSeries, allowed values are

https://github.com/fNIRS/snirf/blob/master/snirf_specification.md#supported-measurementlistkdatatypelabel-values-in-datatimeseries

one should use this to store data analysis software output data - such as those estimated by HOMER or BrainAnalyzIR. If anyone has a sample file and would like to add it to snirf-samples, please share the link, preferable a small sized file.