Richard Dennis
Richard Dennis
If the explosive/implosive models were suitably normalized so that the solutions were stationary, then you might have a shot with global and perturbation methods. For the knife-edge unstable case, even...
On your first point, do I understand correctly that: levels / no r ---- NLsolve works when initialized from steady state with domain [ss*2, ss*0.01] levels / r ---- NLsove...
In general I think that the need for such transforms to keep variables positive is often symtematic of some other problem. In this particular case it's the cause of problems...
I have a multi-threading option for computing the weights, but I haven't tried to use CUDA or any GPU processing; this might be worth looking into in the future. The...
Okay. Thanks for the references.
Thanks for letting me know. I'm happy for ChebyshevApprox to be faster and more useful to more people, so I would be interested to adopting some of your improvements. What's...
Hi Roberto, I have received and read your comment, but I haven't had time to look into it. I believe I know what you are referring to, however, but I'll...
Paul, I have been looking into the issue you raise. Are you able to send me a short model.txt file that illustrates the problem you are having? You are welcome...
Paul, out of interest, have you tried solving that model using a perturbation method? Can you please tell me what domain you are using for the approximation scheme. Posting the...
Roberto and Paul, I have tagged a new release v0.5.8 that I think addresses the two bugs you encountered. Thanks.