Non-Gaussianity amplitude is no longer computed in v0.3 (i.e., in multi-field version). It needs to be reimplemented.
This would require optimizations for InflationValue as right now using expressions large enough as is required for non-Gaussianity would be very slow (because intermediate expressions are recomputed many times and not cached).
Also, non-Gaussianity amplitude should be renamed to be more specific as to which non-Gaussianity it is.
Main issue here is that we need equations for speed of sound and non-Guassianity for a multiple-field general Lagrangian. Otherwise, s-correction to scalar spectral index cannot be computed.
Non-Gaussianity amplitude is no longer computed in v0.3 (i.e., in multi-field version). It needs to be reimplemented.
This would require optimizations for
InflationValue
as right now using expressions large enough as is required for non-Gaussianity would be very slow (because intermediate expressions are recomputed many times and not cached).Also, non-Gaussianity amplitude should be renamed to be more specific as to which non-Gaussianity it is.