def run_smbo(self, max_iters=1000):
global evaluator
# == first things first: load the datamanager
self.reset_data_manager()
# == Initialize SMBO stuff
# first create a scenario
seed = self.seed # TODO
num_params = len(self.config_space.get_hyperparameters())
# allocate a run history
run_history = RunHistory()
meta_runhistory = RunHistory()
meta_runs_dataset_indices = {}
num_run = self.start_num_run
instance_id = self.dataset_name + SENTINEL
# == Train on subset
# before doing anything, let us run the default_cfg
# on a subset of the available data to ensure that
# we at least have some models
# we will try three different ratios of decreasing magnitude
# in the hope that at least on the last one we will be able
# to get a model
n_data = self.datamanager.data['X_train'].shape[0]
subset_ratio = 10000. / n_data
if subset_ratio >= 0.5:
subset_ratio = 0.33
subset_ratios = [subset_ratio, subset_ratio * 0.10]
else:
subset_ratios = [subset_ratio, 500. / n_data]
self.logger.info("Training default configurations on a subset of "
"%d/%d data points." %
(int(n_data * subset_ratio), n_data))
# the time limit for these function evaluations is rigorously
# set to only 1/2 of a full function evaluation
subset_time_limit = max(5, int(self.func_eval_time_limit / 2))
# the configs we want to run on the data subset are:
# 1) the default configs
# 2) a set of configs we selected for training on a subset
subset_configs = [self.config_space.get_default_configuration()] \
+ self.collect_additional_subset_defaults()
subset_config_succesful = [False] * len(subset_configs)
for subset_config_id, next_config in enumerate(subset_configs):
for i, ratio in enumerate(subset_ratios):
self.reset_data_manager()
n_data_subsample = int(n_data * ratio)
# run the config, but throw away the result afterwards
# since this cfg was evaluated only on a subset
# and we don't want to confuse SMAC
self.logger.info(
"Starting to evaluate %d on SUBSET "
"with size %d and time limit %ds.", num_run,
n_data_subsample, subset_time_limit)
self.logger.info(next_config)
_info = eval_with_limits(self.datamanager, self.tmp_dir,
next_config, seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit, subset_time_limit,
n_data_subsample)
(duration, result, _, additional_run_info, status) = _info
self.logger.info(
"Finished evaluating %d. configuration on SUBSET. "
"Duration %f; loss %f; status %s; additional run "
"info: %s ", num_run, duration, result, str(status),
additional_run_info)
num_run += 1
if i < len(subset_ratios) - 1:
if status != StatusType.SUCCESS:
# Do not increase num_run here, because we will try
# the same configuration with less data
self.logger.info(
"A CONFIG did not finish "
" for subset ratio %f -> going smaller", ratio)
continue
else:
self.logger.info(
"Finished SUBSET training sucessfully "
"with ratio %f", ratio)
subset_config_succesful[subset_config_id] = True
break
else:
if status != StatusType.SUCCESS:
self.logger.info(
"A CONFIG did not finish "
" for subset ratio %f.", ratio)
continue
else:
self.logger.info(
"Finished SUBSET training sucessfully "
"with ratio %f", ratio)
subset_config_succesful[subset_config_id] = True
break
# Use the first non-failing configuration from the subsets as the new
# default configuration -> this guards us against the random forest
# failing on large, sparse datasets
default_cfg = None
for subset_config_id, next_config in enumerate(subset_configs):
if subset_config_succesful[subset_config_id]:
default_cfg = next_config
break
if default_cfg is None:
default_cfg = self.config_space.get_default_configuration()
# == METALEARNING suggestions
# we start by evaluating the defaults on the full dataset again
# and add the suggestions from metalearning behind it
if self.metadata_directory is None:
metalearning_directory = os.path.dirname(
autosklearn.metalearning.__file__)
# There is no multilabel data in OpenML
if self.task == MULTILABEL_CLASSIFICATION:
meta_task = BINARY_CLASSIFICATION
else:
meta_task = self.task
metadata_directory = os.path.join(
metalearning_directory, 'files', '%s_%s_%s' %
(METRIC_TO_STRING[self.metric],
TASK_TYPES_TO_STRING[meta_task],
'sparse' if self.datamanager.info['is_sparse'] else 'dense'))
self.metadata_directory = metadata_directory
self.logger.info('Metadata directory: %s', self.metadata_directory)
meta_base = MetaBase(self.config_space, self.metadata_directory)
metafeature_calculation_time_limit = int(self.total_walltime_limit / 4)
metafeature_calculation_start_time = time.time()
meta_features = self._calculate_metafeatures_with_limits(
metafeature_calculation_time_limit)
metafeature_calculation_end_time = time.time()
metafeature_calculation_time_limit = \
metafeature_calculation_time_limit - (
metafeature_calculation_end_time -
metafeature_calculation_start_time)
if metafeature_calculation_time_limit < 1:
self.logger.warning(
'Time limit for metafeature calculation less '
'than 1 seconds (%f). Skipping calculation '
'of metafeatures for encoded dataset.',
metafeature_calculation_time_limit)
meta_features_encoded = None
else:
self.datamanager.perform1HotEncoding()
meta_features_encoded = \
self._calculate_metafeatures_encoded_with_limits(
metafeature_calculation_time_limit)
# In case there is a problem calculating the encoded meta-features
if meta_features is None:
if meta_features_encoded is not None:
meta_features = meta_features_encoded
else:
if meta_features_encoded is not None:
meta_features.metafeature_values.update(
meta_features_encoded.metafeature_values)
if meta_features is not None:
meta_base.add_dataset(instance_id, meta_features)
# Do mean imputation of the meta-features - should be done specific
# for each prediction model!
all_metafeatures = meta_base.get_metafeatures(
features=list(meta_features.keys()))
all_metafeatures.fillna(all_metafeatures.mean(), inplace=True)
metalearning_configurations = self.collect_metalearning_suggestions(
meta_base)
if metalearning_configurations is None:
metalearning_configurations = []
self.reset_data_manager()
self.logger.info('%s', meta_features)
# Convert meta-features into a dictionary because the scenario
# expects a dictionary
meta_features_dict = {}
for dataset, series in all_metafeatures.iterrows():
meta_features_dict[dataset] = series.values
meta_features_list = []
for meta_feature_name in all_metafeatures.columns:
meta_features_list.append(
meta_features[meta_feature_name].value)
meta_features_list = np.array(meta_features_list).reshape((1, -1))
self.logger.info(list(meta_features_dict.keys()))
meta_runs = meta_base.get_all_runs(METRIC_TO_STRING[self.metric])
meta_runs_index = 0
try:
meta_durations = meta_base.get_all_runs('runtime')
read_runtime_data = True
except KeyError:
read_runtime_data = False
self.logger.critical('Cannot read runtime data.')
if self.acquisition_function == 'EIPS':
self.logger.critical(
'Reverting to acquisition function EI!')
self.acquisition_function = 'EI'
for meta_dataset in meta_runs.index:
meta_dataset_start_index = meta_runs_index
for meta_configuration in meta_runs.columns:
if np.isfinite(meta_runs.loc[meta_dataset,
meta_configuration]):
try:
config = meta_base.get_configuration_from_algorithm_index(
meta_configuration)
cost = meta_runs.loc[meta_dataset,
meta_configuration]
if read_runtime_data:
runtime = meta_durations.loc[
meta_dataset, meta_configuration]
else:
runtime = 1
# TODO read out other status types!
meta_runhistory.add(config,
cost,
runtime,
StatusType.SUCCESS,
instance_id=meta_dataset)
meta_runs_index += 1
except:
# TODO maybe add warning
pass
meta_runs_dataset_indices[meta_dataset] = (
meta_dataset_start_index, meta_runs_index)
else:
if self.acquisition_function == 'EIPS':
self.logger.critical('Reverting to acquisition function EI!')
self.acquisition_function = 'EI'
meta_features_list = []
meta_features_dict = {}
metalearning_configurations = []
self.scenario = AutoMLScenario(self.config_space,
self.total_walltime_limit,
self.func_eval_time_limit,
meta_features_dict, self.tmp_dir,
self.shared_mode)
types = get_types(self.config_space, self.scenario.feature_array)
if self.acquisition_function == 'EI':
rh2EPM = RunHistory2EPM4Cost(num_params=num_params,
scenario=self.scenario,
success_states=None,
impute_censored_data=False,
impute_state=None)
model = RandomForestWithInstances(
types,
instance_features=meta_features_list,
seed=1,
num_trees=10)
smac = SMBO(self.scenario, model=model, rng=seed)
elif self.acquisition_function == 'EIPS':
rh2EPM = RunHistory2EPM4EIPS(num_params=num_params,
scenario=self.scenario,
success_states=None,
impute_censored_data=False,
impute_state=None)
model = UncorrelatedMultiObjectiveRandomForestWithInstances(
['cost', 'runtime'],
types,
num_trees=10,
instance_features=meta_features_list,
seed=1)
acquisition_function = EIPS(model)
smac = SMBO(self.scenario,
acquisition_function=acquisition_function,
model=model,
runhistory2epm=rh2EPM,
rng=seed)
else:
raise ValueError('Unknown acquisition function value %s!' %
self.acquisition_function)
# Build a runtime model
# runtime_rf = RandomForestWithInstances(types,
# instance_features=meta_features_list,
# seed=1, num_trees=10)
# runtime_rh2EPM = RunHistory2EPM4EIPS(num_params=num_params,
# scenario=self.scenario,
# success_states=None,
# impute_censored_data=False,
# impute_state=None)
# X_runtime, y_runtime = runtime_rh2EPM.transform(meta_runhistory)
# runtime_rf.train(X_runtime, y_runtime[:, 1].flatten())
X_meta, Y_meta = rh2EPM.transform(meta_runhistory)
# Transform Y_meta on a per-dataset base
for meta_dataset in meta_runs_dataset_indices:
start_index, end_index = meta_runs_dataset_indices[meta_dataset]
end_index += 1 # Python indexing
Y_meta[start_index:end_index, 0]\
[Y_meta[start_index:end_index, 0] >2.0] = 2.0
dataset_minimum = np.min(Y_meta[start_index:end_index, 0])
Y_meta[start_index:end_index,
0] = 1 - ((1. - Y_meta[start_index:end_index, 0]) /
(1. - dataset_minimum))
Y_meta[start_index:end_index, 0]\
[Y_meta[start_index:end_index, 0] > 2] = 2
# == first, evaluate all metelearning and default configurations
for i, next_config in enumerate(
([default_cfg] + metalearning_configurations)):
# Do not evaluate default configurations more than once
if i >= len([default_cfg]) and next_config in [default_cfg]:
continue
config_name = 'meta-learning' if i >= len([default_cfg]) \
else 'default'
self.logger.info(
"Starting to evaluate %d. configuration "
"(%s configuration) with time limit %ds.", num_run,
config_name, self.func_eval_time_limit)
self.logger.info(next_config)
self.reset_data_manager()
info = eval_with_limits(self.datamanager, self.tmp_dir,
next_config, seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit,
self.func_eval_time_limit)
(duration, result, _, additional_run_info, status) = info
run_history.add(config=next_config,
cost=result,
time=duration,
status=status,
instance_id=instance_id,
seed=seed)
run_history.update_cost(next_config, result)
self.logger.info(
"Finished evaluating %d. configuration. "
"Duration %f; loss %f; status %s; additional run "
"info: %s ", num_run, duration, result, str(status),
additional_run_info)
num_run += 1
if smac.incumbent is None:
smac.incumbent = next_config
elif result < run_history.get_cost(smac.incumbent):
smac.incumbent = next_config
if self.scenario.shared_model:
pSMAC.write(run_history=run_history,
output_directory=self.scenario.output_dir,
num_run=self.seed)
# == after metalearning run SMAC loop
smac.runhistory = run_history
smac_iter = 0
finished = False
while not finished:
if self.scenario.shared_model:
pSMAC.read(run_history=run_history,
output_directory=self.scenario.output_dir,
configuration_space=self.config_space,
logger=self.logger)
next_configs = []
time_for_choose_next = -1
try:
X_cfg, Y_cfg = rh2EPM.transform(run_history)
if not run_history.empty():
# Update costs by normalization
dataset_minimum = np.min(Y_cfg[:, 0])
Y_cfg[:, 0] = 1 - ((1. - Y_cfg[:, 0]) /
(1. - dataset_minimum))
Y_cfg[:, 0][Y_cfg[:, 0] > 2] = 2
if len(X_meta) > 0 and len(X_cfg) > 0:
pass
#X_cfg = np.concatenate((X_meta, X_cfg))
#Y_cfg = np.concatenate((Y_meta, Y_cfg))
elif len(X_meta) > 0:
X_cfg = X_meta.copy()
Y_cfg = Y_meta.copy()
elif len(X_cfg) > 0:
X_cfg = X_cfg.copy()
Y_cfg = Y_cfg.copy()
else:
raise ValueError(
'No training data for SMAC random forest!')
self.logger.info('Using %d training points for SMAC.' %
X_cfg.shape[0])
choose_next_start_time = time.time()
next_configs_tmp = smac.choose_next(
X_cfg,
Y_cfg,
num_interleaved_random=110,
num_configurations_by_local_search=10,
num_configurations_by_random_search_sorted=100)
time_for_choose_next = time.time() - choose_next_start_time
self.logger.info('Used %g seconds to find next '
'configurations' % (time_for_choose_next))
next_configs.extend(next_configs_tmp)
# TODO put Exception here!
except Exception as e:
self.logger.error(e)
self.logger.error("Error in getting next configurations "
"with SMAC. Using random configuration!")
next_config = self.config_space.sample_configuration()
next_configs.append(next_config)
models_fitted_this_iteration = 0
start_time_this_iteration = time.time()
for next_config in next_configs:
x_runtime = impute_inactive_values(next_config)
x_runtime = impute_inactive_values(x_runtime).get_array()
# predicted_runtime = runtime_rf.predict_marginalized_over_instances(
# x_runtime.reshape((1, -1)))
# predicted_runtime = np.exp(predicted_runtime[0][0][0]) - 1
self.logger.info(
"Starting to evaluate %d. configuration (from "
"SMAC) with time limit %ds.", num_run,
self.func_eval_time_limit)
self.logger.info(next_config)
self.reset_data_manager()
info = eval_with_limits(self.datamanager, self.tmp_dir,
next_config, seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit,
self.func_eval_time_limit)
(duration, result, _, additional_run_info, status) = info
run_history.add(config=next_config,
cost=result,
time=duration,
status=status,
instance_id=instance_id,
seed=seed)
run_history.update_cost(next_config, result)
#self.logger.info('Predicted runtime %g, true runtime %g',
# predicted_runtime, duration)
# TODO add unittest to make sure everything works fine and
# this does not get outdated!
if smac.incumbent is None:
smac.incumbent = next_config
elif result < run_history.get_cost(smac.incumbent):
smac.incumbent = next_config
self.logger.info(
"Finished evaluating %d. configuration. "
"Duration: %f; loss: %f; status %s; additional "
"run info: %s ", num_run, duration, result, str(status),
additional_run_info)
smac_iter += 1
num_run += 1
models_fitted_this_iteration += 1
time_used_this_iteration = time.time(
) - start_time_this_iteration
if models_fitted_this_iteration >= 2 and \
time_for_choose_next > 0 and \
time_used_this_iteration > time_for_choose_next:
break
elif time_for_choose_next <= 0 and \
models_fitted_this_iteration >= 1:
break
elif models_fitted_this_iteration >= 50:
break
if max_iters is not None:
finished = (smac_iter < max_iters)
if self.scenario.shared_model:
pSMAC.write(run_history=run_history,
output_directory=self.scenario.output_dir,
num_run=self.seed)
class MetaLearningOptimizer(object):
def __init__(self,
dataset_name,
configuration_space,
aslib_directory,
distance='l1',
seed=None,
use_features='',
distance_kwargs=None,
subset='all'):
"""Metalearning optimizer.
Parameters
----------
dataset_name : str
Name of the dataset
configuration_space : HPOlibConfigSpace.configuration_space.ConfigurationSpace
datasets_file : str
Path to an aslib directory
distance : str, "l1" or "l2" or "random"
Distance function to be used by the kNearestDatasets algorithm.
seed
use_features
metric_kwargs
subset
"""
self.dataset_name = dataset_name
self.configuration_space = configuration_space
self.aslib_dir = aslib_directory
self.distance = distance
self.seed = seed
self.use_features = use_features
self.distance_kwargs = distance_kwargs
self.subset = subset
self.kND = None # For caching, makes things faster...
self.meta_base = MetaBase(configuration_space, self.aslib_dir)
self.logger = logging.getLogger(__name__)
def perform_sequential_optimization(self,
target_algorithm=test_function,
time_budget=None,
evaluation_budget=None):
raise NotImplementedError("Right now this is not implemented due to "
"timing issues.")
time_taken = 0
num_evaluations = 0
history = []
self.logger.info("Taking distance measure %s" % self.distance)
while True:
if time_budget is not None and time_taken >= time_budget:
self.logger.info("Reached time budget. Exiting optimization.")
break
if evaluation_budget is not None and \
num_evaluations >= evaluation_budget:
self.logger.info(
"Reached maximum number of evaluations. Exiting "
"optimization.")
break
params = self.metalearning_suggest(history)
fixed_params = OrderedDict()
# Hack to remove all trailing - from the params which are
# accidently in the experiment pickle of the current HPOlib version
for key in params:
if key[0] == "-":
fixed_params[key[1:]] = params[key]
else:
fixed_params[key] = params[key]
self.logger.info(
"%d/%d, parameters: %s" %
(num_evaluations, evaluation_budget, str(fixed_params)))
result = target_algorithm(fixed_params)
history.append(Run(params, result))
num_evaluations += 1
return min([run.result for run in history])
def metalearning_suggest_all(self, exclude_double_configurations=True):
"""Return a list of the best hyperparameters of neighboring datasets"""
# TODO check if _learn was called before!
neighbors = self._learn(exclude_double_configurations)
hp_list = []
for neighbor in neighbors:
try:
configuration = \
self.meta_base.get_configuration_from_algorithm_index(
neighbor[2])
self.logger.info("%s %s %s" %
(neighbor[0], neighbor[1], configuration))
except (KeyError):
self.logger.warning("Configuration %s not found" % neighbor[2])
continue
hp_list.append(configuration)
return hp_list
def metalearning_suggest(self, history):
"""Suggest the next most promising hyperparameters which were not yet evaluated"""
# TODO test the object in the history!
neighbors = self._learn()
# Iterate over all datasets which are sorted ascending by distance
history_with_indices = []
for run in history:
history_with_indices.append(\
self.meta_base.get_algorithm_index_from_configuration(run))
for idx, neighbor in enumerate(neighbors):
already_evaluated = False
# Check if that dataset was already evaluated
for run in history_with_indices:
# If so, return to the outer loop
if neighbor[2] == run:
already_evaluated = True
break
if not already_evaluated:
self.logger.info(
"Nearest dataset with hyperparameters of best value "
"not evaluated yet is %s with a distance of %f" %
(neighbor[0], neighbor[1]))
return self.meta_base.get_configuration_from_algorithm_index(
neighbor[2])
raise StopIteration("No more values available.")
def _learn(self, exclude_double_configurations=True):
dataset_metafeatures, all_other_metafeatures = self._get_metafeatures()
# Remove metafeatures which could not be calculated for the target
# dataset
keep = []
for idx in dataset_metafeatures.index:
if np.isfinite(dataset_metafeatures.loc[idx]):
keep.append(idx)
dataset_metafeatures = dataset_metafeatures.loc[keep]
all_other_metafeatures = all_other_metafeatures.loc[:, keep]
# Do mean imputation of all other metafeatures
all_other_metafeatures = all_other_metafeatures.fillna(
all_other_metafeatures.mean())
if self.kND is None:
# In case that we learn our distance function, get_value the parameters for
# the random forest
if self.distance_kwargs:
rf_params = ast.literal_eval(self.distance_kwargs)
else:
rf_params = None
# To keep the distance the same in every iteration, we create a new
# random state
random_state = sklearn.utils.check_random_state(self.seed)
kND = KNearestDatasets(metric=self.distance,
random_state=random_state,
metric_params=rf_params)
runs = dict()
# TODO move this code to the metabase
for task_id in all_other_metafeatures.index:
try:
runs[task_id] = self.meta_base.get_runs(task_id)
except KeyError:
# TODO should I really except this?
self.logger.warning("Could not find runs for instance %s" %
task_id)
runs[task_id] = pd.Series([], name=task_id)
runs = pd.DataFrame(runs)
kND.fit(all_other_metafeatures, runs)
self.kND = kND
return self.kND.kBestSuggestions(
dataset_metafeatures,
k=-1,
exclude_double_configurations=exclude_double_configurations)
def _get_metafeatures(self):
"""This is inside an extra function for testing purpose"""
# Load the task
self.logger.info("Going to use the metafeature subset: %s",
self.subset)
all_metafeatures = self.meta_base.get_all_metafeatures()
self.logger.info(" ".join(all_metafeatures.columns))
# TODO: buggy and hacky, replace with a list seperated by commas
if self.use_features and \
(type(self.use_features) != str or self.use_features != ''):
#ogger.warn("Going to keep the following features %s",
# str(self.use_features))
if type(self.use_features) == str:
use_features = self.use_features.split(",")
elif type(self.use_features) in (list, np.ndarray):
use_features = self.use_features
else:
raise NotImplementedError(type(self.use_features))
if len(use_features) == 0:
self.logger.info(
"You just tried to remove all metafeatures...")
else:
keep = [
col for col in all_metafeatures.columns
if col in use_features
]
if len(use_features) == 0:
self.logger.info(
"You just tried to remove all metafeatures...")
else:
all_metafeatures = all_metafeatures.loc[:, keep]
self.logger.info(
"Going to keep the following metafeatures:")
self.logger.info(str(keep))
return self._split_metafeature_array(self.dataset_name,
all_metafeatures)
def _split_metafeature_array(self, dataset_name, metafeatures):
"""Split the metafeature array into dataset metafeatures and all other.
This is inside an extra function for testing purpose.
"""
dataset_metafeatures = metafeatures.loc[dataset_name].copy()
metafeatures = metafeatures[metafeatures.index != dataset_name]
return dataset_metafeatures, metafeatures
def read_task_list(self, fh):
dataset_filenames = list()
for line in fh:
line = line.replace("\n", "")
if line:
dataset_filenames.append(line)
else:
raise ValueError("Blank lines in the task list are not "
"supported.")
return dataset_filenames
def read_experiments_list(self, fh):
experiments_list = list()
for line in fh.readlines():
experiments_list.append(line.split())
return experiments_list
class MetaLearningOptimizer(object):
def __init__(self, dataset_name, configuration_space,
aslib_directory, distance='l1', seed=None, use_features='',
distance_kwargs=None, subset='all'):
"""Metalearning optimizer.
Parameters
----------
dataset_name : str
Name of the dataset
configuration_space : HPOlibConfigSpace.configuration_space.ConfigurationSpace
datasets_file : str
Path to an aslib directory
distance : str, "l1" or "l2" or "random"
Distance function to be used by the kNearestDatasets algorithm.
seed
use_features
metric_kwargs
subset
"""
self.dataset_name = dataset_name
self.configuration_space = configuration_space
self.aslib_dir = aslib_directory
self.distance = distance
self.seed = seed
self.use_features = use_features
self.distance_kwargs = distance_kwargs
self.subset = subset
self.kND = None # For caching, makes things faster...
self.meta_base = MetaBase(configuration_space, self.aslib_dir)
self.logger = logging.getLogger(__name__)
def perform_sequential_optimization(self, target_algorithm=test_function,
time_budget=None,
evaluation_budget=None):
raise NotImplementedError("Right now this is not implemented due to "
"timing issues.")
time_taken = 0
num_evaluations = 0
history = []
self.logger.info("Taking distance measure %s" % self.distance)
while True:
if time_budget is not None and time_taken >= time_budget:
self.logger.info("Reached time budget. Exiting optimization.")
break
if evaluation_budget is not None and \
num_evaluations >= evaluation_budget:
self.logger.info("Reached maximum number of evaluations. Exiting "
"optimization.")
break
params = self.metalearning_suggest(history)
fixed_params = OrderedDict()
# Hack to remove all trailing - from the params which are
# accidently in the experiment pickle of the current HPOlib version
for key in params:
if key[0] == "-":
fixed_params[key[1:]] = params[key]
else:
fixed_params[key] = params[key]
self.logger.info("%d/%d, parameters: %s" % (num_evaluations,
evaluation_budget,
str(fixed_params)))
result = target_algorithm(fixed_params)
history.append(Run(params, result))
num_evaluations += 1
return min([run.result for run in history])
def metalearning_suggest_all(self, exclude_double_configurations=True):
"""Return a list of the best hyperparameters of neighboring datasets"""
# TODO check if _learn was called before!
neighbors = self._learn(exclude_double_configurations)
hp_list = []
for neighbor in neighbors:
try:
configuration = \
self.meta_base.get_configuration_from_algorithm_index(
neighbor[2])
self.logger.info("%s %s %s" % (neighbor[0], neighbor[1], configuration))
except (KeyError):
self.logger.warning("Configuration %s not found" % neighbor[2])
continue
hp_list.append(configuration)
return hp_list
def metalearning_suggest(self, history):
"""Suggest the next most promising hyperparameters which were not yet evaluated"""
# TODO test the object in the history!
neighbors = self._learn()
# Iterate over all datasets which are sorted ascending by distance
history_with_indices = []
for run in history:
history_with_indices.append(\
self.meta_base.get_algorithm_index_from_configuration(run))
for idx, neighbor in enumerate(neighbors):
already_evaluated = False
# Check if that dataset was already evaluated
for run in history_with_indices:
# If so, return to the outer loop
if neighbor[2] == run:
already_evaluated = True
break
if not already_evaluated:
self.logger.info("Nearest dataset with hyperparameters of best value "
"not evaluated yet is %s with a distance of %f" %
(neighbor[0], neighbor[1]))
return self.meta_base.get_configuration_from_algorithm_index(
neighbor[2])
raise StopIteration("No more values available.")
def _learn(self, exclude_double_configurations=True):
dataset_metafeatures, all_other_metafeatures = self._get_metafeatures()
# Remove metafeatures which could not be calculated for the target
# dataset
keep = []
for idx in dataset_metafeatures.index:
if np.isfinite(dataset_metafeatures.loc[idx]):
keep.append(idx)
dataset_metafeatures = dataset_metafeatures.loc[keep]
all_other_metafeatures = all_other_metafeatures.loc[:,keep]
# Do mean imputation of all other metafeatures
all_other_metafeatures = all_other_metafeatures.fillna(
all_other_metafeatures.mean())
if self.kND is None:
# In case that we learn our distance function, get_value the parameters for
# the random forest
if self.distance_kwargs:
rf_params = ast.literal_eval(self.distance_kwargs)
else:
rf_params = None
# To keep the distance the same in every iteration, we create a new
# random state
random_state = sklearn.utils.check_random_state(self.seed)
kND = KNearestDatasets(metric=self.distance,
random_state=random_state,
metric_params=rf_params)
runs = dict()
# TODO move this code to the metabase
for task_id in all_other_metafeatures.index:
try:
runs[task_id] = self.meta_base.get_runs(task_id)
except KeyError:
# TODO should I really except this?
self.logger.warning("Could not find runs for instance %s" % task_id)
runs[task_id] = pd.Series([], name=task_id)
runs = pd.DataFrame(runs)
kND.fit(all_other_metafeatures, runs)
self.kND = kND
return self.kND.kBestSuggestions(dataset_metafeatures, k=-1,
exclude_double_configurations=exclude_double_configurations)
def _get_metafeatures(self):
"""This is inside an extra function for testing purpose"""
# Load the task
self.logger.info("Going to use the metafeature subset: %s", self.subset)
all_metafeatures = self.meta_base.get_all_metafeatures()
self.logger.info(" ".join(all_metafeatures.columns))
# TODO: buggy and hacky, replace with a list seperated by commas
if self.use_features and \
(type(self.use_features) != str or self.use_features != ''):
#ogger.warn("Going to keep the following features %s",
# str(self.use_features))
if type(self.use_features) == str:
use_features = self.use_features.split(",")
elif type(self.use_features) in (list, np.ndarray):
use_features = self.use_features
else:
raise NotImplementedError(type(self.use_features))
if len(use_features) == 0:
self.logger.info("You just tried to remove all metafeatures...")
else:
keep = [col for col in all_metafeatures.columns if col in use_features]
if len(use_features) == 0:
self.logger.info("You just tried to remove all metafeatures...")
else:
all_metafeatures = all_metafeatures.loc[:,keep]
self.logger.info("Going to keep the following metafeatures:")
self.logger.info(str(keep))
return self._split_metafeature_array(self.dataset_name, all_metafeatures)
def _split_metafeature_array(self, dataset_name, metafeatures):
"""Split the metafeature array into dataset metafeatures and all other.
This is inside an extra function for testing purpose.
"""
dataset_metafeatures = metafeatures.loc[dataset_name].copy()
metafeatures = metafeatures[metafeatures.index != dataset_name]
return dataset_metafeatures, metafeatures
def read_task_list(self, fh):
dataset_filenames = list()
for line in fh:
line = line.replace("\n", "")
if line:
dataset_filenames.append(line)
else:
raise ValueError("Blank lines in the task list are not "
"supported.")
return dataset_filenames
def read_experiments_list(self, fh):
experiments_list = list()
for line in fh.readlines():
experiments_list.append(line.split())
return experiments_list
def run_smbo(self, max_iters=1000):
global evaluator
# == first things first: load the datamanager
self.reset_data_manager()
# == Initialize SMBO stuff
# first create a scenario
seed = self.seed # TODO
num_params = len(self.config_space.get_hyperparameters())
# allocate a run history
run_history = RunHistory()
meta_runhistory = RunHistory()
meta_runs_dataset_indices = {}
num_run = self.start_num_run
instance_id = self.dataset_name + SENTINEL
# == Train on subset
# before doing anything, let us run the default_cfg
# on a subset of the available data to ensure that
# we at least have some models
# we will try three different ratios of decreasing magnitude
# in the hope that at least on the last one we will be able
# to get a model
n_data = self.datamanager.data['X_train'].shape[0]
subset_ratio = 10000. / n_data
if subset_ratio >= 0.5:
subset_ratio = 0.33
subset_ratios = [subset_ratio, subset_ratio * 0.10]
else:
subset_ratios = [subset_ratio, 500. / n_data]
self.logger.info("Training default configurations on a subset of "
"%d/%d data points." %
(int(n_data * subset_ratio), n_data))
# the time limit for these function evaluations is rigorously
# set to only 1/2 of a full function evaluation
subset_time_limit = max(5, int(self.func_eval_time_limit / 2))
# the configs we want to run on the data subset are:
# 1) the default configs
# 2) a set of configs we selected for training on a subset
subset_configs = [self.config_space.get_default_configuration()] \
+ self.collect_additional_subset_defaults()
subset_config_succesful = [False] * len(subset_configs)
for subset_config_id, next_config in enumerate(subset_configs):
for i, ratio in enumerate(subset_ratios):
self.reset_data_manager()
n_data_subsample = int(n_data * ratio)
# run the config, but throw away the result afterwards
# since this cfg was evaluated only on a subset
# and we don't want to confuse SMAC
self.logger.info("Starting to evaluate %d on SUBSET "
"with size %d and time limit %ds.",
num_run, n_data_subsample,
subset_time_limit)
self.logger.info(next_config)
_info = eval_with_limits(
self.datamanager, self.tmp_dir, next_config,
seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit,
subset_time_limit, n_data_subsample)
(duration, result, _, additional_run_info, status) = _info
self.logger.info("Finished evaluating %d. configuration on SUBSET. "
"Duration %f; loss %f; status %s; additional run "
"info: %s ", num_run, duration, result,
str(status), additional_run_info)
num_run += 1
if i < len(subset_ratios) - 1:
if status != StatusType.SUCCESS:
# Do not increase num_run here, because we will try
# the same configuration with less data
self.logger.info("A CONFIG did not finish "
" for subset ratio %f -> going smaller",
ratio)
continue
else:
self.logger.info("Finished SUBSET training sucessfully "
"with ratio %f", ratio)
subset_config_succesful[subset_config_id] = True
break
else:
if status != StatusType.SUCCESS:
self.logger.info("A CONFIG did not finish "
" for subset ratio %f.",
ratio)
continue
else:
self.logger.info("Finished SUBSET training sucessfully "
"with ratio %f", ratio)
subset_config_succesful[subset_config_id] = True
break
# Use the first non-failing configuration from the subsets as the new
# default configuration -> this guards us against the random forest
# failing on large, sparse datasets
default_cfg = None
for subset_config_id, next_config in enumerate(subset_configs):
if subset_config_succesful[subset_config_id]:
default_cfg = next_config
break
if default_cfg is None:
default_cfg = self.config_space.get_default_configuration()
# == METALEARNING suggestions
# we start by evaluating the defaults on the full dataset again
# and add the suggestions from metalearning behind it
if self.metadata_directory is None:
metalearning_directory = os.path.dirname(
autosklearn.metalearning.__file__)
# There is no multilabel data in OpenML
if self.task == MULTILABEL_CLASSIFICATION:
meta_task = BINARY_CLASSIFICATION
else:
meta_task = self.task
metadata_directory = os.path.join(
metalearning_directory, 'files',
'%s_%s_%s' % (METRIC_TO_STRING[self.metric],
TASK_TYPES_TO_STRING[meta_task],
'sparse' if self.datamanager.info['is_sparse']
else 'dense'))
self.metadata_directory = metadata_directory
self.logger.info('Metadata directory: %s', self.metadata_directory)
meta_base = MetaBase(self.config_space, self.metadata_directory)
metafeature_calculation_time_limit = int(
self.total_walltime_limit / 4)
metafeature_calculation_start_time = time.time()
meta_features = self._calculate_metafeatures_with_limits(
metafeature_calculation_time_limit)
metafeature_calculation_end_time = time.time()
metafeature_calculation_time_limit = \
metafeature_calculation_time_limit - (
metafeature_calculation_end_time -
metafeature_calculation_start_time)
if metafeature_calculation_time_limit < 1:
self.logger.warning('Time limit for metafeature calculation less '
'than 1 seconds (%f). Skipping calculation '
'of metafeatures for encoded dataset.',
metafeature_calculation_time_limit)
meta_features_encoded = None
else:
self.datamanager.perform1HotEncoding()
meta_features_encoded = \
self._calculate_metafeatures_encoded_with_limits(
metafeature_calculation_time_limit)
# In case there is a problem calculating the encoded meta-features
if meta_features is None:
if meta_features_encoded is not None:
meta_features = meta_features_encoded
else:
if meta_features_encoded is not None:
meta_features.metafeature_values.update(
meta_features_encoded.metafeature_values)
if meta_features is not None:
meta_base.add_dataset(instance_id, meta_features)
# Do mean imputation of the meta-features - should be done specific
# for each prediction model!
all_metafeatures = meta_base.get_metafeatures(
features=list(meta_features.keys()))
all_metafeatures.fillna(all_metafeatures.mean(), inplace=True)
metalearning_configurations = self.collect_metalearning_suggestions(
meta_base)
if metalearning_configurations is None:
metalearning_configurations = []
self.reset_data_manager()
self.logger.info('%s', meta_features)
# Convert meta-features into a dictionary because the scenario
# expects a dictionary
meta_features_dict = {}
for dataset, series in all_metafeatures.iterrows():
meta_features_dict[dataset] = series.values
meta_features_list = []
for meta_feature_name in all_metafeatures.columns:
meta_features_list.append(meta_features[meta_feature_name].value)
meta_features_list = np.array(meta_features_list).reshape((1, -1))
self.logger.info(list(meta_features_dict.keys()))
meta_runs = meta_base.get_all_runs(METRIC_TO_STRING[self.metric])
meta_runs_index = 0
try:
meta_durations = meta_base.get_all_runs('runtime')
read_runtime_data = True
except KeyError:
read_runtime_data = False
self.logger.critical('Cannot read runtime data.')
if self.acquisition_function == 'EIPS':
self.logger.critical('Reverting to acquisition function EI!')
self.acquisition_function = 'EI'
for meta_dataset in meta_runs.index:
meta_dataset_start_index = meta_runs_index
for meta_configuration in meta_runs.columns:
if np.isfinite(meta_runs.loc[meta_dataset, meta_configuration]):
try:
config = meta_base.get_configuration_from_algorithm_index(
meta_configuration)
cost = meta_runs.loc[meta_dataset, meta_configuration]
if read_runtime_data:
runtime = meta_durations.loc[meta_dataset,
meta_configuration]
else:
runtime = 1
# TODO read out other status types!
meta_runhistory.add(config, cost, runtime,
StatusType.SUCCESS,
instance_id=meta_dataset)
meta_runs_index += 1
except:
# TODO maybe add warning
pass
meta_runs_dataset_indices[meta_dataset] = (
meta_dataset_start_index, meta_runs_index)
else:
if self.acquisition_function == 'EIPS':
self.logger.critical('Reverting to acquisition function EI!')
self.acquisition_function = 'EI'
meta_features_list = []
meta_features_dict = {}
metalearning_configurations = []
self.scenario = AutoMLScenario(self.config_space,
self.total_walltime_limit,
self.func_eval_time_limit,
meta_features_dict,
self.tmp_dir,
self.shared_mode)
types = get_types(self.config_space, self.scenario.feature_array)
if self.acquisition_function == 'EI':
rh2EPM = RunHistory2EPM4Cost(num_params=num_params,
scenario=self.scenario,
success_states=None,
impute_censored_data=False,
impute_state=None)
model = RandomForestWithInstances(types,
instance_features=meta_features_list,
seed=1, num_trees=10)
smac = SMBO(self.scenario, model=model,
rng=seed)
elif self.acquisition_function == 'EIPS':
rh2EPM = RunHistory2EPM4EIPS(num_params=num_params,
scenario=self.scenario,
success_states=None,
impute_censored_data=False,
impute_state=None)
model = UncorrelatedMultiObjectiveRandomForestWithInstances(
['cost', 'runtime'], types, num_trees = 10,
instance_features=meta_features_list, seed=1)
acquisition_function = EIPS(model)
smac = SMBO(self.scenario,
acquisition_function=acquisition_function,
model=model, runhistory2epm=rh2EPM, rng=seed)
else:
raise ValueError('Unknown acquisition function value %s!' %
self.acquisition_function)
# Build a runtime model
# runtime_rf = RandomForestWithInstances(types,
# instance_features=meta_features_list,
# seed=1, num_trees=10)
# runtime_rh2EPM = RunHistory2EPM4EIPS(num_params=num_params,
# scenario=self.scenario,
# success_states=None,
# impute_censored_data=False,
# impute_state=None)
# X_runtime, y_runtime = runtime_rh2EPM.transform(meta_runhistory)
# runtime_rf.train(X_runtime, y_runtime[:, 1].flatten())
X_meta, Y_meta = rh2EPM.transform(meta_runhistory)
# Transform Y_meta on a per-dataset base
for meta_dataset in meta_runs_dataset_indices:
start_index, end_index = meta_runs_dataset_indices[meta_dataset]
end_index += 1 # Python indexing
Y_meta[start_index:end_index, 0]\
[Y_meta[start_index:end_index, 0] >2.0] = 2.0
dataset_minimum = np.min(Y_meta[start_index:end_index, 0])
Y_meta[start_index:end_index, 0] = 1 - (
(1. - Y_meta[start_index:end_index, 0]) /
(1. - dataset_minimum))
Y_meta[start_index:end_index, 0]\
[Y_meta[start_index:end_index, 0] > 2] = 2
# == first, evaluate all metelearning and default configurations
for i, next_config in enumerate(([default_cfg] +
metalearning_configurations)):
# Do not evaluate default configurations more than once
if i >= len([default_cfg]) and next_config in [default_cfg]:
continue
config_name = 'meta-learning' if i >= len([default_cfg]) \
else 'default'
self.logger.info("Starting to evaluate %d. configuration "
"(%s configuration) with time limit %ds.",
num_run, config_name, self.func_eval_time_limit)
self.logger.info(next_config)
self.reset_data_manager()
info = eval_with_limits(self.datamanager, self.tmp_dir, next_config,
seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit,
self.func_eval_time_limit)
(duration, result, _, additional_run_info, status) = info
run_history.add(config=next_config, cost=result,
time=duration , status=status,
instance_id=instance_id, seed=seed)
run_history.update_cost(next_config, result)
self.logger.info("Finished evaluating %d. configuration. "
"Duration %f; loss %f; status %s; additional run "
"info: %s ", num_run, duration, result,
str(status), additional_run_info)
num_run += 1
if smac.incumbent is None:
smac.incumbent = next_config
elif result < run_history.get_cost(smac.incumbent):
smac.incumbent = next_config
if self.scenario.shared_model:
pSMAC.write(run_history=run_history,
output_directory=self.scenario.output_dir,
num_run=self.seed)
# == after metalearning run SMAC loop
smac.runhistory = run_history
smac_iter = 0
finished = False
while not finished:
if self.scenario.shared_model:
pSMAC.read(run_history=run_history,
output_directory=self.scenario.output_dir,
configuration_space=self.config_space,
logger=self.logger)
next_configs = []
time_for_choose_next = -1
try:
X_cfg, Y_cfg = rh2EPM.transform(run_history)
if not run_history.empty():
# Update costs by normalization
dataset_minimum = np.min(Y_cfg[:, 0])
Y_cfg[:, 0] = 1 - ((1. - Y_cfg[:, 0]) /
(1. - dataset_minimum))
Y_cfg[:, 0][Y_cfg[:, 0] > 2] = 2
if len(X_meta) > 0 and len(X_cfg) > 0:
pass
#X_cfg = np.concatenate((X_meta, X_cfg))
#Y_cfg = np.concatenate((Y_meta, Y_cfg))
elif len(X_meta) > 0:
X_cfg = X_meta.copy()
Y_cfg = Y_meta.copy()
elif len(X_cfg) > 0:
X_cfg = X_cfg.copy()
Y_cfg = Y_cfg.copy()
else:
raise ValueError('No training data for SMAC random forest!')
self.logger.info('Using %d training points for SMAC.' %
X_cfg.shape[0])
choose_next_start_time = time.time()
next_configs_tmp = smac.choose_next(X_cfg, Y_cfg,
num_interleaved_random=110,
num_configurations_by_local_search=10,
num_configurations_by_random_search_sorted=100)
time_for_choose_next = time.time() - choose_next_start_time
self.logger.info('Used %g seconds to find next '
'configurations' % (time_for_choose_next))
next_configs.extend(next_configs_tmp)
# TODO put Exception here!
except Exception as e:
self.logger.error(e)
self.logger.error("Error in getting next configurations "
"with SMAC. Using random configuration!")
next_config = self.config_space.sample_configuration()
next_configs.append(next_config)
models_fitted_this_iteration = 0
start_time_this_iteration = time.time()
for next_config in next_configs:
x_runtime = impute_inactive_values(next_config)
x_runtime = impute_inactive_values(x_runtime).get_array()
# predicted_runtime = runtime_rf.predict_marginalized_over_instances(
# x_runtime.reshape((1, -1)))
# predicted_runtime = np.exp(predicted_runtime[0][0][0]) - 1
self.logger.info("Starting to evaluate %d. configuration (from "
"SMAC) with time limit %ds.", num_run,
self.func_eval_time_limit)
self.logger.info(next_config)
self.reset_data_manager()
info = eval_with_limits(self.datamanager, self.tmp_dir, next_config,
seed, num_run,
self.resampling_strategy,
self.resampling_strategy_args,
self.memory_limit,
self.func_eval_time_limit)
(duration, result, _, additional_run_info, status) = info
run_history.add(config=next_config, cost=result,
time=duration , status=status,
instance_id=instance_id, seed=seed)
run_history.update_cost(next_config, result)
#self.logger.info('Predicted runtime %g, true runtime %g',
# predicted_runtime, duration)
# TODO add unittest to make sure everything works fine and
# this does not get outdated!
if smac.incumbent is None:
smac.incumbent = next_config
elif result < run_history.get_cost(smac.incumbent):
smac.incumbent = next_config
self.logger.info("Finished evaluating %d. configuration. "
"Duration: %f; loss: %f; status %s; additional "
"run info: %s ", num_run, duration, result,
str(status), additional_run_info)
smac_iter += 1
num_run += 1
models_fitted_this_iteration += 1
time_used_this_iteration = time.time() - start_time_this_iteration
if models_fitted_this_iteration >= 2 and \
time_for_choose_next > 0 and \
time_used_this_iteration > time_for_choose_next:
break
elif time_for_choose_next <= 0 and \
models_fitted_this_iteration >= 1:
break
elif models_fitted_this_iteration >= 50:
break
if max_iters is not None:
finished = (smac_iter < max_iters)
if self.scenario.shared_model:
pSMAC.write(run_history=run_history,
output_directory=self.scenario.output_dir,
num_run=self.seed)
