def fit(self, pipeline_config, run_result_dir):
instance_file_config_parser = ConfigFileParser([
ConfigOption(name='path', type='directory', required=True),
ConfigOption(name='is_classification', type=to_bool,
required=True),
ConfigOption(name='is_multilabel', type=to_bool, required=True),
ConfigOption(name='num_features', type=int, required=True),
ConfigOption(name='categorical_features',
type=bool,
required=True,
list=True),
ConfigOption(
name='instance_shape',
type=[ast.literal_eval, lambda x: isinstance(x, tuple)],
required=True)
])
instance_info = instance_file_config_parser.read(
os.path.join(run_result_dir, 'instance.info'))
instance_info = instance_file_config_parser.set_defaults(instance_info)
dm = DataManager()
if instance_info["is_multilabel"]:
dm.problem_type = ProblemType.FeatureMultilabel
elif instance_info["is_classification"]:
dm.problem_type = ProblemType.FeatureClassification
else:
dm.problem_type = ProblemType.FeatureClassification
return {'instance_info': instance_info, 'data_manager': dm}
def get_pipeline_config(self,
throw_error_if_invalid=True,
**pipeline_config):
"""Get the full pipeline config given a partial pipeline config
Keyword Arguments:
throw_error_if_invalid {bool} -- Throw an error if invalid config option is defined (default: {True})
Returns:
dict -- the full config for the pipeline, containing values for all options
"""
options = self.get_pipeline_config_options()
conditions = self.get_pipeline_config_conditions()
parser = ConfigFileParser(options)
pipeline_config = parser.set_defaults(
pipeline_config, throw_error_if_invalid=throw_error_if_invalid)
# check the conditions e.g. max_budget > min_budget
for c in conditions:
try:
c(pipeline_config)
except Exception as e:
if throw_error_if_invalid:
raise
print(e)
traceback.print_exc()
return pipeline_config
def get_pipeline_config(self,
throw_error_if_invalid=True,
**pipeline_config):
options = self.get_pipeline_config_options()
conditions = self.get_pipeline_config_conditions()
parser = ConfigFileParser(options)
pipeline_config = parser.set_defaults(
pipeline_config, throw_error_if_invalid=throw_error_if_invalid)
for c in conditions:
try:
c(pipeline_config)
except Exception as e:
if throw_error_if_invalid:
raise
print(e)
traceback.print_exc()
return pipeline_config
def get_pipeline_config_options(self):
options = [
ConfigOption(
"default_dataset_download_dir",
default=ConfigFileParser.get_autonet_home(),
type='directory',
info="Directory default datasets will be downloaded to."),
ConfigOption("dataloader_worker", default=1, type=int),
ConfigOption("dataloader_cache_size_mb", default=0, type=int)
]
return options
def get_pipeline_config_options(self):
options = [
ConfigOption("instances",
default=None,
type='directory',
required=True),
ConfigOption("instance_slice", default=None, type=str),
ConfigOption("dataset_root",
default=ConfigFileParser.get_autonet_home(),
type='directory'),
]
return options
def read_data(self, path, y):
if path is None:
return None, None, None
if not isinstance(path, str):
return path, y, str(path)[0:300]
if not os.path.isabs(path):
path = os.path.abspath(os.path.join(ConfigFileParser.get_autonet_home(), path))
if not os.path.exists(path):
raise ValueError('Path ' + str(path) + ' is not a valid path.')
im = ImageManager()
im.read_data(path, is_classification=True)
return im.X_train, im.Y_train, path
def get_pipeline_config_options(self):
options = [
ConfigOption(name="file_extensions",
default=['.png', '.jpg', '.JPEG', '.pgm'],
type=str,
list=True),
ConfigOption(
name="images_shape",
default=[3, 32, 32],
type=int,
list=True,
info=
"Image size input to the networks, images will be rescaled to this."
),
ConfigOption(
name="images_root_folders",
default=[ConfigFileParser.get_autonet_home()],
type='directory',
list=True,
info="Directory relative to which image paths are given."),
ConfigOption(name="max_class_size", default=None, type=int),
]
return options
def get_autonet_config_file_parser(self):
return ConfigFileParser(self.pipeline.get_pipeline_config_options())
parser.add_argument("--host_config", default=None, help="Override some configs according to host specifics.")
parser.add_argument('benchmark', help='The benchmark to run')
args = parser.parse_args()
assert "-" not in args.run_id, "The run id must not contain a minus"
benchmark_config_file = args.benchmark
host_config_file = args.host_config
benchmark = Benchmark()
config_parser = benchmark.get_benchmark_config_file_parser()
benchmark_config = config_parser.read(benchmark_config_file)
benchmark_config.update(config_parser.read(host_config_file))
if (args.result_dir is not None):
benchmark_config['result_dir'] = os.path.join(ConfigFileParser.get_autonet_home(), args.result_dir)
if (args.partial_benchmark is not None):
split = args.partial_benchmark.split(',')
if (len(split) > 0):
benchmark_config['instance_slice'] = split[0]
if (len(split) > 1):
benchmark_config['autonet_config_slice'] = split[1]
if (len(split) > 2):
benchmark_config['run_number_range'] = split[2]
benchmark_config['run_id'] = args.run_id
benchmark_config['task_id'] = args.task_id
benchmark.run_benchmark(**benchmark_config)
def get_benchmark_config_file_parser(self):
return ConfigFileParser(
self.benchmark_pipeline.get_pipeline_config_options())
if "-" in args.run_id_range:
run_id_range = range(int(args.run_id_range.split("-")[0]),
int(args.run_id_range.split("-")[1]) + 1)
else:
run_id_range = range(int(args.run_id_range),
int(args.run_id_range) + 1)
benchmark_config_file = args.benchmark
host_config_file = args.host_config
benchmark = Benchmark()
config_parser = benchmark.get_benchmark_config_file_parser()
benchmark_config = config_parser.read(benchmark_config_file)
benchmark_config.update(config_parser.read(host_config_file))
if (args.result_dir is not None):
benchmark_config['result_dir'] = os.path.join(
ConfigFileParser.get_autonet_home(), args.result_dir)
benchmark_config['run_id_range'] = run_id_range
benchmark_config['plot_logs'] = args.plot_logs.split(
",") if args.plot_logs is not None else list()
benchmark_config['only_finished_runs'] = args.only_finished_runs
benchmark_config['output_folder'] = args.output_folder
benchmark_config['scale_uncertainty'] = args.scale_uncertainty
benchmark_config['agglomeration'] = args.agglomeration
benchmark_config['font_size'] = args.font_size
benchmark.visualize_benchmark(**benchmark_config)
default=[],
nargs="+",
type=str,
help="Additional args specified in template")
parser.add_argument(
"runscript", help="The script template used to submit job on cluster.")
parser.add_argument('benchmark', help='The benchmark to run')
args = parser.parse_args()
# parse the runscript template
with open(args.runscript, "r") as f:
runscript_template = list(f)
runscript_name = os.path.basename(
args.runscript if not args.runscript.endswith(".template") else args.
runscript[:-9])
autonet_home = ConfigFileParser.get_autonet_home()
host_config_orig = [
l[13:] for l in runscript_template if l.startswith("#HOST_CONFIG ")
][0].strip()
host_config_file = os.path.join(
autonet_home, host_config_orig
) if not os.path.isabs(host_config_orig) else host_config_orig
# parse template args
runscript_template_args = [
l[19:].strip().split() for l in runscript_template
if l.startswith("#TEMPLATE_ARGUMENT ")
]
parsed_template_args = dict()
for variable_name, default in runscript_template_args:
try:
