While doing some small migrations to ruff as part of #77, I noticed that get_loss/cost/learning_curve() are always done on a ConfigResult object, i.e. get_loss(config_result.result, ...). I will just move these to be methods directly on the config object instead of round-abouting to the aforementioned standalone functions. This means any consumers of ConfigResult can just do config_result.loss(...). This is also hiding behind one more abstraction in the BaseOptimizer which essentially wraps these methods with the extra parameters ... filled in.
While doing some small migrations to ruff as part of #77, I noticed that
get_loss/cost/learning_curve()are always done on aConfigResultobject, i.e.get_loss(config_result.result, ...). I will just move these to be methods directly on the config object instead of round-abouting to the aforementioned standalone functions. This means any consumers ofConfigResultcan just doconfig_result.loss(...). This is also hiding behind one more abstraction in theBaseOptimizerwhich essentially wraps these methods with the extra parameters...filled in.