[FIX] Fix example misalignment

[DimitriPapadopoulos]

Partially fixes #154. This fixes the example. However I believe a more “theoretical” discussion is required, explaining that raw data are stored directly under the dataset root folder, the same way source data are stored in subfolder sourcedata and derived data in subfolder derivatives.

[Remi-Gau]

This looks good to me. Might avoid some confusion indeed.

[Remi-Gau]

I had made a figure (when this whole thing was on google doc) to help explain how raw / derivatives / source relate to one another. It is now sitting on a PR / issue on the bids starter kit repo.

I think that I had in mind the idea of putting in the BIDS specs but never got around to to it.

I think @chrisfilo had made some comments on the figure being too "specific" and might lead people to assume that the content of source and derivative had to the one described in the figure.

I am happy to change it if there is interest in adding this to the specs.

[DimitriPapadopoulos]

Indeed I cannot find in the standard a clear statement about the location of raw data. As a result colleagues of mine believe you may put raw data in a subfolder of the root folder such as rawdata. On the other hand I do see that source data belong to sourcedata and derived data to derivatives. Fixing the example should clarify a little bit.

I would also like to add a sourcedata subfolder to the example to complete it. Would you agree to adding sourcedata?

Perhaps we don't need to show the contents of the subfolder though. Should we show the contents of sourcedata and other subfolders as in the above figure? I am all for it. This is an example after all!

Finally what about adding bow-drawing characters to this example - like the tree command? Something like this:

├── sub-control01
│   ├── anat
.    .
.    .
.    .
├── code
│   └──  deface.py
├── derivatives
├── README
├── CHANGES
├── participants.tsv
└── dataset_description.json

GitHub doesn't render it nicely but perhaps Read the Docs does. Perhaps a real figure like in #105 is preferable.

[chrisgorgo]

Let's merge this as is since it clearly solves an existing problem (alignment).

I totally agree that we need more clarification on where source vs raw data should go. @Remi-Gau I think your plots were very close and maybe we should bring them back. I also like the idea of listing sourcedata and derivatives folders without contents.

@DimitriPapadopoulos consider also sending a PR adding yourself to https://github.com/bids-standard/bids-specification/blob/master/src/99-appendices/01-contributors.md and https://github.com/bids-standard/bids-specification/blob/master/CODEOWNERS