for testing and deploying your application
for finding and fixing issues
for empowering human code reviews
from ._result import Result
class ObjectiveAdapter:
"""Maps *pos* → (score, metrics, params)."""
def __init__(self, conv, objective):
self._conv = conv
self._objective = objective # user callable
def _call_objective(self, pos):
"""Run the underlying objective and normalise outputs."""
params = self._conv.value2para(self._conv.position2value(pos))
out = self._objective(params)
if isinstance(out, tuple):
score, metrics = out
else:
score, metrics = float(out), {}
result = Result(score, metrics)
return result, params
# keep one public entry point so subclasses can override cleanly
def __call__(self, pos):
return self._call_objective(pos)
SimonBlanke /
Gradient-Free-Optimizers
