V1 batched

[BalzaniEdoardo]

V1 tutorial

I have modified the v1 tutorial in order to fit a spatiotemporal receptive field with 10ms resolution using the stochastic optimization approach.

Changes from previous version

What we had before was inferring the spatial receptive field with an STA and using that to project the stimulus into a single dimension. The temporal component was fit via a GLM. This allowed a fine temporal resolution; In the new version, the full spatiotemporal filter is computed using basis to reduce the feature dimensionality. A stochastic approach to parameter learning was employed to show how one can avoid to build the full design matrix (that likely would be very large).

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 97.30%. Comparing base (87d3c29) to head (68cffb9).

Additional details and impacted files

@@             Coverage Diff              @@
##           development     #153   +/-   ##
============================================
  Coverage        97.30%   97.30%           
============================================
  Files               15       15           
  Lines             1485     1485           
============================================
  Hits              1445     1445           
  Misses              40       40           

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.