MODEL: Adjust cell spacing

[asoplata]

cc: @ntolley @carolinafernandezp @katduecker

I don't know the details, but as Nick mentioned in the last dev meeting, there is a scientific issue having to do with placement of the cells in space. IIRC, the issue with the placement of the inhibitory and excitatory cells should not, in principle, affect the current-dipole output of our simulations, but it will likely affect any LFP/CSD output. In order for us to also work towards scientifically meaningful LFP/CSD, it sounded like this is an issue that needs to be addressed. I've made this issue for us to track our work.

[asoplata]

Irresponsible suggestion: it may make sense to also deal with #342 and #568 in the course of solving this issue.

[carolinafernandezp]

I can work on this. The problem is that the same-layer basket cells and pyramidal neurons should be on the same plane. Currently, the basket cells are arbitrarily placed slightly above (L5) and slightly below (L2) their respective pyramidal cell layers.

This is happening in _calc_basket_coord where zdiff (the depth of L2 relative to L5 pyramidal cell somas) gets offset by a weight of 0.8 and 0.2 for the L2 and L5 basket cells, respectively (which ends up being 261.48 μm since zdiff is hard-coded to 1307.4 μm).

I'm rerunning the LFP/CSD simulations to see how they are affected by removing the offset.