def init(self):
if not self.data_exist():
self.download()
src_dir = path.join(datasets_dir(self.root_dir), 'download',
'training2017')
shutil.move(src_dir,
path.join(datasets_dir(self.root_dir), self.name()))
print(
f'Successfully downloaded dataset {self.name()} to {self.root_dir}.'
)
if self.features_exist():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir),
f'{self.name()}Features.pkl'))
return
if self.dataset_exists():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir), f'{self.name()}.pkl'))
else:
self.data = pd.read_csv(path.join(datasets_dir(self.root_dir),
self.name(), 'REFERENCE.csv'),
names=['Record', 'Label'],
dtype={'Label': 'category'})
self.data['Fs'] = 300
def init(self):
if not self.data_exist():
self.download()
for dir in ['TrainingSet1', 'TrainingSet2', 'TrainingSet3']:
src_dir = path.join(datasets_dir(self.root_dir), 'download',
dir)
shutil.copytree(src_dir,
path.join(datasets_dir(self.root_dir),
self.name()),
dirs_exist_ok=True,
copy_function=shutil.move)
os.rmdir(src_dir)
print(
f'Successfully downloaded dataset {self.name()} to {self.root_dir}.'
)
if self.features_exist():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir),
f'{self.name()}Features.pkl'))
return
if self.dataset_exists():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir), f'{self.name()}.pkl'))
else:
self.data = pd.read_csv(path.join(datasets_dir(self.root_dir),
self.name(), 'REFERENCE.csv'),
dtype={
'First_label': 'Int64',
'Second_label': 'Int64',
'Third_label': 'Int64'
})
self.data.columns = ['Record', 'Label', 'Label2', 'Label3']
self.data = self.data[self.data['Label2'].isnull()].reset_index(
drop=True)
self.data['Fs'] = 500
def init(self):
if self.features_exist():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir),
f'{self.name()}Features.pkl'))
return
if self.dataset_exists():
self.data = pd.read_pickle(
path.join(data_dir(self.root_dir), f'{self.name()}.pkl'))
else:
self.data = pd.read_pickle(
path.join(datasets_dir(self.root_dir), self.name(),
'private_dataset.pkl')).drop(columns=['Signal'])
self.data.columns = ['Device', 'Label', 'Record']
self.data['Fs'] = 200
def download(self):
"""
Downloads and extracts the dataset's resources from self.resources.
"""
if self.data_exist():
return
download_dir = path.join(datasets_dir(self.root_dir), 'download')
ensure_directory_exists(download_dir)
for filename, (url, md5) in self.resources.items():
download_and_extract_archive(url,
download_root=download_dir,
filename=filename,
md5=md5,
remove_finished=False)
print('Downloading done!')
def data_exist(self):
"""
:return: True if the dataset is extracted in the datasets folder.
"""
return path.exists(
path.join(datasets_dir(self.root_dir), f'{self.name()}'))
def __init__(self, root_dir='.', transform=None, data=None, split=None):
super().__init__(root_dir, transform, data, split)
self.signals = pd.read_pickle(
path.join(datasets_dir(self.root_dir), self.name(),
'private_dataset.pkl'))
def read_record(self, record):
mat = loadmat(
path.join(datasets_dir(self.root_dir), self.name(), record))
signal = mat['ECG'][0, 0][2][self.lead]
return signal.astype(np.float32)