Skip to content

Conversation

@spalato
Copy link
Contributor

@spalato spalato commented Apr 8, 2025

Added regularization=0.0 argument to RadialBasis type, constructor and to _calc_coeffs. The regularization parameter is added on the diagonal of the D matrix to avoid SingularException.

Checklist

  • Appropriate tests were added. See test\Radials.jl
  • Any code changes were done in a way that does not break public API. Handled by existing tests in test\Radials.jl
  • All documentation related to code changes were updated
  • The new code follows the
    contributor guidelines, in particular the SciML Style Guide and
    COLPRAC.
  • Any new documentation only uses public API

Additional context

Optimizing with a RadialBasis surrogate often result in many points close to the optimum being added. This results in a SingularValueException, preventing successful optimization. Addition of a small regularization parameter of the diagonal of the D matrix prevents this issue. The value is added as an optional argument with default value of 0, in order to preserve default behavior.

See issue #482.

spalato added 2 commits April 8, 2025 11:27
Added `regularization=0.0` argument to `RadialBasis` type, constructor and to `_calc_coeffs`. The regulatization parameter is added on the diagnal of the D matrix to avoid `SingularException`.
@sathvikbhagavan
Copy link
Member

@ChrisRackauckas, can you trigger CI?

@ChrisRackauckas ChrisRackauckas merged commit a86ebf4 into SciML:master Apr 10, 2025
6 of 9 checks passed
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

Successfully merging this pull request may close these issues.

3 participants