Skip to content
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

Disable stop criteria (x tolerance) do not work for Sbplx #452

Open
sgherrero opened this issue Apr 14, 2022 · 5 comments
Open

Disable stop criteria (x tolerance) do not work for Sbplx #452

sgherrero opened this issue Apr 14, 2022 · 5 comments

Comments

@sgherrero
Copy link

Hello,
I am using the algorithm Sbplx (LN_SBPLX). I have disabled the stop criteria based on the x tolerance (relative and absolute) using the (C++) methods set_xtol_rel(value) and set_xtol_abs(value) with a negative value. Indeed, I have read that the stop criterias are disabled by default but I have explicitely disabled. However, the algorithm stops and it returns the code NLOPT_XTOL_REACHED = 4. Could you expain me this behaviour?
A reason could be that the Subplex is necessary this stop criteria but I have not found it in the documentation.
Thanks in advance for the help.

Sara

@UnixJunkie
Copy link

I think I observe the same problem with auglag (global) parameterized by sbplx (local).
Optimization stops with NLOPT_FTOL_REACHED while I have called both

  Nlopt.set_ftol_abs local (-1.0)
  Nlopt.set_ftol_abs global (-1.0)

@UnixJunkie
Copy link

I am using libnlopt-dev version 2.7.1-3build1 on Ubuntu Linux.
I wonder if this is a bug affecting several of the optimization algorithms.

@jschueller
Copy link
Collaborator

jschueller commented Jul 6, 2023

can you provide a minimal reproducer example please ?

@UnixJunkie
Copy link

UnixJunkie commented Jul 6, 2023

By default, those values are all 0s:

15:34:05.32|I: local_ftol_abs before: 0.000000
15:34:05.32|I: local_ftol_rel before: 0.000000
15:34:05.32|I: global_ftol_abs before: 0.000000
15:34:05.32|I: global_ftol_rel before: 0.000000

but the optimization still stops w/ NLOPT_FTOL_REACHED.

@UnixJunkie
Copy link

I am using the OCaml interface and working on a large unpublished program, the behavior I observe is not easy to replicate/minimize.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants