Closed ffreyer closed 2 years ago
Calculating this only takes about 9ns, but since it's calculated in propose_local! it's called very frequently. Throwing this in a RefValue{Float64} in model (Ref{Float64} is abstract) may actually reduce the total runtime by a few percent.
propose_local!
RefValue{Float64}
Ref{Float64}
Did some quick and dirty benchmarking:
propose_local
But the simulation spent <1% in propose_local, so there isn't really a point in optimizing it further.
Calculating this only takes about 9ns, but since it's calculated in
propose_local!
it's called very frequently. Throwing this in aRefValue{Float64}
in model (Ref{Float64}
is abstract) may actually reduce the total runtime by a few percent.