def test_File_delimiter(self) -> None:
with self.subTest(format='rfc'):
with csv.File(self.rfc_path) as file:
self.assertEqual(file.delimiter, self.rfc_sep)
with self.subTest(format='rlang'):
with csv.File(self.rlang_path) as file:
self.assertEqual(file.delimiter, self.rlang_sep)
def test_File_name(self) -> None:
with self.subTest(format='rfc'):
with csv.File(self.rfc_path) as file:
self.assertEqual(file.name, self.rfc_path.name)
with self.subTest(format='rlang'):
with csv.File(self.rlang_path) as file:
self.assertEqual(file.name, self.rlang_path.name)
def test_File_rownames(self) -> None:
with self.subTest(format='rfc'):
with csv.File(self.rfc_path) as file:
self.assertEqual(file.rownames, None)
with csv.File(self.rfc_path, namecol='name') as file:
self.assertEqual(file.rownames, self.rownames)
with self.subTest(format='rlang'):
with csv.File(self.rlang_path) as file:
self.assertEqual(file.rownames, self.rownames)
def test_File_init(self) -> None:
with self.subTest(format='rfc'):
with csv.File(
self.rfc_path, header=self.rfc_header, comment=self.comment,
delimiter=self.rfc_sep) as file:
self.assertIsInstance(file, csv.File)
with self.subTest(format='rlang'):
with csv.File(
self.rlang_path, header=self.rlang_header, comment=self.comment,
delimiter=self.rlang_sep) as file:
self.assertIsInstance(file, csv.File)
def load(self, path):
"""Get dataset configuration and dataset tables.
Args:
path (string): csv file containing dataset configuration and
dataset table.
"""
# Get instance of CSV-file
file = csv.File(path)
# Get configuration from CSV comment lines
comment = file.comment
scheme = {
'name': str,
'branch': str,
'version': int,
'about': str,
'author': str,
'email': str,
'license': str,
'filetype': str,
'application': str,
'preprocessing': dict,
'type': str,
'labelformat': str
}
config = ini.decode(comment, flat=True, scheme=scheme)
if 'name' in config:
name = config['name']
else:
name = env.basename(path)
config['name'] = name
if 'type' not in config:
config['type'] = 'base.Dataset'
# Add column and row filters
config['colfilter'] = {'*': ['*:*']}
config['rowfilter'] = {'*': ['*:*'], name: [name + ':*']}
# Load data
names = list(file.header)
names[0] = 'label'
data = array.from_tuples(file.read(), names=tuple(names))
config['table'] = {name: config.copy()}
config['table'][name]['fraction'] = 1.0
config['columns'] = tuple()
config['colmapping'] = {}
config['table'][name]['columns'] = []
for column in data.dtype.names:
if column == 'label':
continue
config['columns'] += (('', column), )
config['colmapping'][column] = column
config['table'][name]['columns'].append(column)
# Get data table from CSV data
tables = {name: data}
return {'config': config, 'tables': tables}