Rémi Lafage

Hi @mariaelenavalencia! You should not have exactly the same error on Windows as above as it is triggered by `xcrun` which is specific to MacOS install (see the last lines)....

FWIW, I would like to mention here that [external solvers](https://github.com/argmin-rs/argmin#external-solvers-compatible-with-argmin) handle constraints. In both case, cost function returns a vector: the first component is the actual cost value, the remaining...

No, there is no native support for parallel computing or GPU.

Hi, You're right, currently there is no API for sampling MFK GPs or accessing covariance matrices directly. If you want to implement such methods take a look at [krg_sampling.py](https://github.com/SMTorg/smt/blob/8e1b7ea2f8971831a259a877df2d0a30fb448b53/smt/utils/krg_sampling.py) where...

Hi. First what is the dimension of your training samples? Your surrogates may suffer from overfitting as you experience degradation with number of samples. If you have numerous samples, you...

Did you try the `intel-mkl-static` backend? On Windows, it works for me with: ``` ndarray = { version = "0.15", features = ["blas"] } ndarray-linalg = { version = "0.15",...

Hi. Thank you for reporting. I think we've got an issue here. How do you call `KPLS()`, what options are you using? What is the dimension of your training inputs...

1) the point of using `KPLS` or `KPLSK` is to choose `n_comp < num_params` to get actual dimension reduction otherwise you'd better use `KRG`. What is the value of `num_params`?...

Hi. No there is no native implementation of save/load for kriging-based models in SMT (see also [the doc](https://smt.readthedocs.io/en/stable/_src_docs/surrogate_models.html#how-to-save-and-load-trained-surrogate-models)). But I guess you already know that 😉 which leads to your...

From #76, I understand that download from S3 is obsolete. Downstream crates have to upgrade to intel-mkl-src 0.8. @termoshtt correct?