Saving of membrane potentials is not working

[torbjone]

For certain cases, it might make sense to save the membrane potentials of the cells, and for example in the "example_brunel.py" file, "somav" is already in "savelist", which would indicate that it should be saved. This does however not seem to be working.

A hack that seemed to work locally was to add "recordSingleContribFrac=1," as an argument to the Population class in "example_brunel.py" (L334): https://github.com/INM-6/hybridLFPy/blob/aec00dc019a5de8d5de1e07dfb1c4a9e23122463/examples/example_brunel.py#L334 and add "somav" to the list on line 755 in "population.py": https://github.com/INM-6/hybridLFPy/blob/aec00dc019a5de8d5de1e07dfb1c4a9e23122463/hybridLFPy/population.py#L755

However, this will also save all single-cell extracellular potentials (and possible CSDs), which might be a bit overkill for large simulations. We should probably at least add "somav" to the list on line 755 in "population.py", or implement a new way to save somav without also saving all single cell LFP/CSD contributions.

Any thoughts?

[espenhgn]

Hi, Perhaps this is better fixed when the LFPy@2.2.dev0 branch is patched into the hybridLFPy@nest3 branch. Then savelist should be for LFPy.Cell attributes exclusively and predictions made using LFPykit.model instances would be summed up and saved per population and in total by default.

But if you need a quick fix, perhaps modify the Population.collect_data and Postprocess.run() method accordingly.

[espenhgn]

Closing this issue - this should now work in the current nest3 branch since merging #29. Note that this working branch requires LFPy@2.2.dev0 and NEST@master (tested with https://github.com/nest/nest-simulator/commit/24de43dc21c568e017839eeb335253c2bc2d487d) as outlined in this new Dockerfile recipe: https://github.com/INM-6/hybridLFPy/blob/nest3/examples/Hagen_et_al_2016_cercor/Dockerfile

[espenhgn]

Should add: savelist output should end up in the files $SCRATCH/<simulation_output_folder>/populations/<Y>_savelist.h5