debug

from deepchem.utils.save import log

from deepchem.hyper import HyperparamOpt

from deepchem.molnet.run_benchmark_models import benchmark_classification, benchmark_regression

from deepchem.utils.dependencies import pyGPGO_covfunc, pyGPGO_acquisition, \

    pyGPGO_surrogates_GaussianProcess, pyGPGO_GPGO

class GaussianProcessHyperparamOpt(HyperparamOpt):

  """

        try:

          train_scores, valid_scores, _ = benchmark_classification(

              train_dataset, valid_dataset, None, ['task_placeholder']*n_tasks,

              output_transformers, n_features, metric, model, 

              output_transformers, n_features, metric, self.model_class, 

              hyper_parameters=hyper_parameters)

        except AssertionError:

          train_scores, valid_scores, _ = benchmark_regression(

              train_dataset, valid_dataset, None, ['task_placeholder']*n_tasks,

              output_transformers, n_features, metric, model, 

              output_transformers, n_features, metric, self.model_class, 

              hyper_parameters=hyper_parameters)

        return valid_scores[model][metric[0].name]

        return valid_scores[self.model_class][metric[0].name]

      else:

        model_dir = tempfile.mkdtemp()

        model = self.model_class(hyper_parameters, model_dir)