def from_pandas(df, cache_dir):
if df.size == 0:
return SessionDataInfo()
fcn_local_path = lambda rec: Path(cache_dir).joinpath( # noqa
rec['lab'],
'public',
'projects',
rec['project'],
'ALF',
rec['subject'],
rec['start_time'].isoformat()[:10],
#fcn_local_path = lambda rec: Path(cache_dir).joinpath( # noqa
# rec['lab'], 'Subjects', rec['subject'], rec['start_time'].isoformat()[:10],
str(rec['number']).zfill(3),
rec['collection'],
rec['name'])
nrecs = df.shape[0]
return SessionDataInfo(
dataset_type=df.dataset_type.to_list(),
dataset_id=list(parquet.np2str(df[['id_0', 'id_1']])),
local_path=df.apply(fcn_local_path, axis=1).to_list(),
eid=list(parquet.np2str(df[['eid_0', 'eid_1']])),
url=[None for _ in range(nrecs)],
data=[None for _ in range(nrecs)],
hash=df.hash.to_list(),
file_size=df.file_size.to_list())
def test_uuids_conversions(self):
str_uuid = 'a3df91c8-52a6-4afa-957b-3479a7d0897c'
one_np_uuid = np.array([-411333541468446813, 8973933150224022421])
two_np_uuid = np.tile(one_np_uuid, [2, 1])
# array gives a list
self.assertTrue(all(map(lambda x: x == str_uuid, np2str(two_np_uuid))))
# single uuid gives a string
self.assertTrue(np2str(one_np_uuid) == str_uuid)
def test_uuids_conversions(self):
str_uuid = 'a3df91c8-52a6-4afa-957b-3479a7d0897c'
one_np_uuid = np.array([-411333541468446813, 8973933150224022421])
two_np_uuid = np.tile(one_np_uuid, [2, 1])
# array gives a list
self.assertTrue(all(map(lambda x: x == str_uuid, np2str(two_np_uuid))))
# single uuid gives a string
self.assertTrue(np2str(one_np_uuid) == str_uuid)
# list uuids with some None entries
uuid_list = [
'bc74f49f33ec0f7545ebc03f0490bdf6',
'c5779e6d02ae6d1d6772df40a1a94243', None,
'643371c81724378d34e04a60ef8769f4'
]
assert np.all(str2np(uuid_list)[2, :] == 0)
def eid_from_path(self, path_obj):
"""
From a local path, gets the experiment id
:param path_obj: local path or list of local paths
:return: eid or list of eids
"""
# If path_obj is a list recurse through it and return a list
if isinstance(path_obj, list):
path_obj = [Path(x) for x in path_obj]
eid_list = []
for p in path_obj:
eid_list.append(self.eid_from_path(p))
return eid_list
# else ensure the path ends with mouse,date, number
path_obj = Path(path_obj)
session_path = alfio.get_session_path(path_obj)
# if path does not have a date and a number, or cache is empty return None
if session_path is None or self._cache.size == 0:
return None
# fetch eid from cache
ind = ((self._cache['subject'] == session_path.parts[-3]) &
(self._cache['start_time'].apply(
lambda x: x.isoformat()[:10] == session_path.parts[-2])) &
(self._cache['number']) == int(session_path.parts[-1]))
ind = np.where(ind.to_numpy())[0]
if ind.size > 0:
return parquet.np2str(self._cache[['eid_0', 'eid_1']].iloc[ind[0]])
def eid_from_path(self, path_obj, use_cache=True):
"""
From a local path, gets the experiment id
:param path_obj: local path or list of local paths
:param use_cache: if set to False, will force database connection
:return: eid or list of eids
"""
# If path_obj is a list recurse through it and return a list
if isinstance(path_obj, list):
path_obj = [Path(x) for x in path_obj]
eid_list = []
for p in path_obj:
eid_list.append(self.eid_from_path(p))
return eid_list
# else ensure the path ends with mouse,date, number
path_obj = Path(path_obj)
session_path = alfio.get_session_path(path_obj)
# if path does not have a date and a number return None
if session_path is None:
return None
# try the cached info to possibly avoid hitting database
if self._cache.size > 0 and use_cache:
ind = ((self._cache['subject'] == session_path.parts[-3]) &
(self._cache['start_time'].apply(
lambda x: x.isoformat()[:10] == session_path.parts[-2]))
& (self._cache['number']) == int(session_path.parts[-1]))
ind = np.where(ind.to_numpy())[0]
if ind.size > 0:
return parquet.np2str(self._cache[['eid_0',
'eid_1']].iloc[ind[0]])
# if not search for subj, date, number XXX: hits the DB
uuid = self.search(subjects=session_path.parts[-3],
date_range=session_path.parts[-2],
number=session_path.parts[-1])
# Return the uuid if any
return uuid[0] if uuid else None