Forgotten Exponential?

[olynch]

I feel like there should be an exp here, am I missing something?

AlgebraicPetri.jl/src/AlgebraicPetri.jl

Line 137 in 8c66422

du[sname(pn, j)] = sum(log_rates[i] * dt[i,j] for i in 1:nt(pn))

Ah, the real problem is that they should no longer be called log_rates.

[jpfairbanks]

Yeah, we tried to solve in log-space for stability, but ran into problems and reversed course. Any input on making this code more stable would be well received.

[olynch]

This wouldn't necessarily make this more stable, but would we get a speedup if we code-genned this, to remove all allocations and loops?

[jpfairbanks]

That would be useful as an additional functionality. If you have one big Petri net with many parameters, you want to compile it. If you are doing model selection over many small nets, you want to interpret them.

[mehalter]

Resolved with #107