-
Notifications
You must be signed in to change notification settings - Fork 92
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Variable Step ODE Solver ProgressMeter #15
Comments
ProgressMeter only works when you know how many iterations it's going to take; if that's not true, then you'll have to have some other way of displaying progress. Optim.jl optionally lets you set I'm going to close this, since there really isn't anything that needs to change in ProgressMeter. |
Actually, for a lot of cases, ODE progress reporting would be quite trivial if ProgressMeter supported floating-point steps (and not just unit steps from 1 to n iterations, as today), since one often knows the start and end time of the iterval one integrates over. It would have to be coupled with some effort in ODE.jl (specifically, something like SciML/ODE.jl#27 would be needed, so we could do things between integration steps) but one could envision something like
where ProgressMeter re-scales itself to e.g. 0-100%, and a call to Doing this probably means quite a re-write of ProgressMeter, so I don't expect it will be done anytime soon unless @timholy finds a use case for it himself, but I'm sure a PR with this functionality would be well received 😄 |
Good point. Yes, I'd be happy to consider a pull request. |
Was able to get things going with Sundials and ProgressMeter. Needed the sleep, as my test function was a very simple one for ProgressMeter to show any response. function testfunc(len)
p = ProgressMeter.Progress(len,0.01)
for k = 2:len
flag = Sundials.CVode(cvode_mem, tout[k], y, t, Sundials.CV_NORMAL)
yres[k,:] = y
sleep(0.0001)
ProgressMeter.next!(p)
end
end
println("Testing original interface...")
testfunc(length(tout)) Thanks @timholy for the progressmeter |
Can variable step ode solvers be linked with ProgressMeter.jl to get a somewhat equivalent of Matlab's
odeprint
Sundials progress monitoring, any such stuffs?
The text was updated successfully, but these errors were encountered: