def load_pyradigms(dataset_paths, sub_group=None):
"""Reads in a list of datasets in pyradigm format.
Parameters
----------
dataset_paths : iterable
List of paths to pyradigm dataset
sub_group : iterable
subset of classes to return. Default: return all classes.
If sub_group is specified, returns only that subset of classes for all datasets.
Raises
------
ValueError
If all the datasets do not contain the request subset of classes.
"""
if sub_group is not None:
sub_group = set(sub_group)
# loading datasets
datasets = list()
for fp in dataset_paths:
if not pexists(fp):
raise IOError("Dataset @ {} does not exist.".format(fp))
try:
# there is an internal validation of dataset
ds_in = MLDataset(fp)
except:
print("Dataset @ {} is not a valid MLDataset!".format(fp))
raise
class_set = set(ds_in.class_set)
if sub_group is None or sub_group == class_set:
ds_out = ds_in
elif sub_group < class_set: # < on sets is an issubset operation
ds_out = ds_in.get_class(sub_group)
else:
raise ValueError(
'One or more classes in {} does not exist in\n{}'.format(
sub_group, fp))
# add the valid dataset to list
datasets.append(ds_out)
return datasets
test_dataset.description = 'test dataset'
print(test_dataset)
print('default format:\n {}'.format(test_dataset))
print('full repr :\n {:full}'.format(test_dataset))
print('string/short :\n {:s}'.format(test_dataset))
class_set, label_set, class_sizes = test_dataset.summarize_classes()
reloaded_dataset = MLDataset(filepath=out_file,
description='reloaded test_dataset')
copy_dataset = MLDataset(in_dataset=test_dataset)
rand_index = np.random.randint(0, len(class_set), 1)[0]
random_class_name = class_set[rand_index]
random_class_ds = test_dataset.get_class(random_class_name)
other_classes_ds = test_dataset - random_class_ds
other_class_set = set(class_set) - set([random_class_name])
other_classes_get_with_list = test_dataset.get_class(other_class_set)
recombined = other_classes_ds + random_class_ds
empty_dataset = MLDataset()
test2 = MLDataset()
test3 = MLDataset()
# TODO write tests for CLI
test_dataset.description = 'test dataset'
print(test_dataset)
print('default format:\n {}'.format(test_dataset))
print('full repr :\n {:full}'.format(test_dataset))
print('string/short :\n {:s}'.format(test_dataset))
class_set, label_set, class_sizes = test_dataset.summarize_classes()
reloaded_dataset = MLDataset(filepath=out_file, description='reloaded test_dataset')
copy_dataset = MLDataset(in_dataset=test_dataset)
rand_index = np.random.randint(0,len(class_set),1)[0]
random_class_name = class_set[rand_index]
random_class_ds = test_dataset.get_class(random_class_name)
other_classes_ds = test_dataset - random_class_ds
other_class_set = set(class_set)-set([random_class_name])
other_classes_get_with_list = test_dataset.get_class(other_class_set)
recombined = other_classes_ds + random_class_ds
empty_dataset = MLDataset()
test2 = MLDataset()
test3 = MLDataset()
# TODO write tests for CLI