[Feature] Cells data extraction about cellfinder HOT 3 CLOSED

So that's all possible (apart from radius, as there's no way to extract that with cellfinder).

So would something like this work (in both sample and standard space)?

Or would you rather them all in one csv?

What kind of analysis/visualisation are you planning on? I'm working on another tool, cellfinder-analyst which will incorporate group-level analyses.

from cellfinder.

I think that having it all in one csv would be the easier.

It's not so much for anything I had planned, but more something that would be useful in general. This is the kind of output that one would use for any kind of analysis/visualisation outside of cellfinder so it'd be good to have an option to generate this data.

For instance in a cluttered scene one might want to look at cells in one brain region at the time (e.g. to see in which subdivision of the region are located).

As an example here in brainreder I just visualise the cells in the isocortex:

cells = pd.read_hdf(cellsfile)

# ... find in which structure each cell is and add it to the data frame ...

# Select only cells in the isocortex
regions = scene.get_structure_descendants('Isocortex').acronym
cells = cells.loc[cells.region.isin(regions)]

But to do this I had to determined in which region each cell is from its coordinates.
This is fairly easy and fast to do in BrainRender but it'd be even easier to just have it in a .csv or .h5 file.

from cellfinder.

Took nearly a year, but this is added in 476ccb1.

There is now a file saved by default (cellfinder_output/analysis/all_points.csv) which looks something like:

There's no hierarchy because not every BrainGlobe atlas will necessarily have a hierarchy (and I'm lazy).

from cellfinder.