Michael Zingale comments

Results 336 comments of


                                            Michael Zingale

other 3-alpha rates?

Note: JINA ReacLib use: Title Revised rates for the stellar triple-a process from measurement of 12C nuclear resonances Publication Nature 433, 136-139 Year 2005 Experimental Yes Author H.O.U. Fynbo et...

allow retry with different Jacobian

ah indeed, so maybe the first step is to just do that refactor

allow retry with different Jacobian

my read of this is that is NFLAG < 0, then we had a failure (singular, or convergence failure), and we should redo the Jacobian, perhaps switching to numerical from...

allow retry with different Jacobian

Ah, okay, I understand now. Only `ICF` makes it out of `dvnlsd`, so indeed, if we force a change of Jacobian type in `dvjac` when `ICF > 0`, then I...

allow retry with different Jacobian

if you drop this on a branch I'll run a test with and without it this evening.

allow retry with different Jacobian

we should also consider the idea of not integrating T on retry

what if we pass VODE failure message out of burn through burn_t

okay, that's better than what we have now, so I'll look into that

create a temp burn_state for NSE bailout

blocked by #744

test_eos does not work for all networks

do you have an idea in mind on how to generalize this? I can just make the composition evenly distributed through the different species.

[WIP] Switch Jacobian type on a previous nonlinear solve failure

the idea behind this PR is that if the numerical Jacobian is not working well, we should attempt the analytic Jacobian. This seems to work well for the Castro subchandra...