def _run_ensemble_builder(tmp_dir,
output_dir,
basename,
time_for_task,
task,
metric,
ensemble_size,
ensemble_nbest,
ensemble_indices_dir,
watcher,
log_function):
task_name = 'runEnsemble'
watcher.start_task(task_name)
time_left_for_ensembles = max(0, time_for_task - watcher.wall_elapsed(
basename))
log_function(
'Start Ensemble with %5.2fsec time left' % time_left_for_ensembles)
proc_ensembles = submit_process.run_ensemble_builder(
tmp_dir=tmp_dir,
dataset_name=basename,
task_type=task,
metric=metric,
limit=time_left_for_ensembles,
output_dir=output_dir,
ensemble_size=ensemble_size,
ensemble_nbest=ensemble_nbest,
seed=get_auto_seed(),
ensemble_indices_output_dir=ensemble_indices_dir
)
watcher.stop_task(task_name)
return proc_ensembles
def run_ensemble_builder(self, time_left_for_ensembles=None, max_iterations=None, ensemble_size=None):
if self._ensemble_size > 0 or ensemble_size is not None:
task_name = "runEnsemble"
self._stopwatch.start_task(task_name)
if time_left_for_ensembles is None:
time_left_for_ensembles = max(0, self._time_for_task - self._stopwatch.wall_elapsed(self._dataset_name))
if max_iterations is None:
max_iterations = -1
if ensemble_size is None:
ensemble_size = self._ensemble_size
# It can happen that run_ensemble_builder is called without
# calling fit.
if self._logger:
self._logger.info("Start Ensemble with %5.2fsec time left" % time_left_for_ensembles)
proc_ensembles = submit_process.run_ensemble_builder(
tmp_dir=self._tmp_dir,
dataset_name=self._dataset_name,
task_type=self._task,
metric=self._metric,
limit=time_left_for_ensembles,
output_dir=self._output_dir,
ensemble_size=ensemble_size,
ensemble_nbest=self._ensemble_nbest,
seed=self._seed,
shared_mode=self._shared_mode,
max_iterations=max_iterations,
precision=self.precision,
)
self._stopwatch.stop_task(task_name)
return proc_ensembles
else:
self._logger.info("Not starting ensemble script due to ensemble " "size 0.")
return None
def run_ensemble_builder(self,
time_left_for_ensembles=None,
max_iterations=None,
ensemble_size=None):
if self._ensemble_size > 0 or ensemble_size is not None:
task_name = 'runEnsemble'
self._stopwatch.start_task(task_name)
if time_left_for_ensembles is None:
time_left_for_ensembles = max(0,
self._time_for_task - self._stopwatch.wall_elapsed(
self._dataset_name))
if max_iterations is None:
max_iterations = -1
if ensemble_size is None:
ensemble_size = self._ensemble_size
# It can happen that run_ensemble_builder is called without
# calling fit.
if self._logger:
self._logger.info(
'Start Ensemble with %5.2fsec time left' % time_left_for_ensembles)
proc_ensembles = submit_process.run_ensemble_builder(
tmp_dir=self._tmp_dir,
dataset_name=self._dataset_name,
task_type=self._task,
metric=self._metric,
limit=time_left_for_ensembles,
output_dir=self._output_dir,
ensemble_size=ensemble_size,
ensemble_nbest=self._ensemble_nbest,
seed=self._seed,
shared_mode=self._shared_mode,
max_iterations=max_iterations,
precision=self.precision
)
self._stopwatch.stop_task(task_name)
return proc_ensembles
else:
self._logger.info('Not starting ensemble script due to ensemble '
'size 0.')
return None
