Source code for smt_optim.utils.get_fmin

import numpy as np

from smt_optim.utils.constraints import compute_rscv




def get_fmin(
    f: np.ndarray,
    c: np.ndarray | None = None,
    c_type: list[str] | None = None,
    rscv_tol: float = 0.0,
) -> float:

    if c is None:
        return min(f)

    else:
        if c_type is None:
            c_type = ["less" for _ in range(c.shape[1])]

        rscv = compute_rscv(c, c_type)

        feasible_mask = np.where(rscv <= rscv_tol, True, False)

        if np.any(feasible_mask):
            return f[feasible_mask].min()
        else:
            idx = np.argmin(rscv)
            return f[idx]