JuliaSmoothOptimizers / ADNLPModels.jl

Other
29 stars 12 forks source link

Inaccurate display #215

Closed dpo closed 3 months ago

dpo commented 4 months ago
julia> nls_opt = ADNLSModel(resid, ones(5), 202, jtprod_backend = ADNLPModels.ZygoteADJtprod)                                                                                                                            
ADNLSModel - Nonlinear least-squares model with automatic differentiation backend ADModelBackend{                                                                                                                        
  ForwardDiffADGradient,                                                                                                                                                                                                 
  ForwardDiffADHvprod,                                                                                                                                                                                                   
  EmptyADbackend,
  EmptyADbackend,
  EmptyADbackend,
  ForwardDiffADHessian,
  EmptyADbackend,
}
…

The non-default AD backend for jtprod is not reflected in the output.

tmigot commented 3 months ago

Hi @dpo ! jtprod_backend is for the constraint Jacobian transpose vector product. If you plan to use it for the residual Jacobian transpose then it is.

Nevertheless, the show wasn't defined to NLS backend... #216 should fix it.

dpo commented 3 months ago

Of course! Thank you!