def test_timevar_ncom_hour_values_dap(self):
datafile = "http://edac-dap3.northerngulfinstitute.org/thredds/dodsC/US_East/ncom_relo_useast_u_2011080200/ncom_relo_useast_u_2011080200_t072.nc"
# Manually extract
# time:units = hour since 2000-01-01 00:00:00
dt = datetime(2000, 1, 1, tzinfo=pytz.utc)
ds = netCDF4.Dataset(datafile)
data = ds.variables['time'][:]
# Convert to days
factor = 24
data = data / factor
# Timevar extract
tvar = Timevar(datafile, name='time', tzinfo=pytz.utc)
assert np.allclose(data, tvar.days)
# Now compare datetimes
data = data.tolist()
jds = []
for x in data:
jds.append(dt + timedelta(days=x))
assert (jds == tvar.dates).all()
ds.close()
def test_timevar_roms_seconds_values(self):
datafile = os.path.join(data_path, "ocean_avg_synoptic_seg22.nc")
# Manually extract
# ocean_time:units = "seconds since 1990-01-01 00:00:00" ;
# ocean_time:calendar = "gregorian" ;
dt = datetime(1990, 1, 1, tzinfo=pytz.utc)
ds = netCDF4.Dataset(datafile)
data = ds.variables['ocean_time'][:]
# Convert to days
factor = 60 * 60 * 24
data = data / factor
# Timevar extract
tvar = Timevar(datafile, name='ocean_time')
assert np.allclose(data, tvar.days)
# Now compare datetimes
data = data.tolist()
jds = []
for x in data:
jds.append(dt + timedelta(days=x))
assert (jds == tvar.dates).all()
ds.close()
def test_timevar_length(self):
datafile = os.path.join(data_path, "pws_L2_2012040100.nc")
# Manually extract
ds = netCDF4.Dataset(datafile)
data = ds.variables['time'][:]
# Timevar extract
tvar = Timevar(datafile)
assert data.shape == tvar.shape
datafile = os.path.join(data_path, "ocean_avg_synoptic_seg22.nc")
# Manually extract
ds = netCDF4.Dataset(datafile)
data = ds.variables['ocean_time'][:]
# Timevar extract
tvar = Timevar(datafile, name='ocean_time')
assert data.shape == tvar.shape
ds.close()
def test_timevar_pickle(self):
# NOTE: pickle always worked same process - multiprocessing or a second process had the trouble.
# still, we override pickle methods so this will catch any regressions
datafile = os.path.join(data_path, "marcooshfradar20120331.nc")
tvar = Timevar(datafile, name='time', tzinfo=pytz.utc)
with tempfile.TemporaryFile() as temp:
pickle.dump(tvar, temp)
temp.seek(0)
t2var = pickle.load(temp)
assert np.allclose(tvar, t2var)
assert hasattr(t2var, '_units')
assert tvar._units == t2var._units
def gettimevar(self, var=None, use_cache=True):
#return self._timevar
assert var in self._current_variables
timevar = None
if use_cache is True:
if self._checkcache(var):
timevar = self._coordcache[var].time
else:
self._coordcache[var] = cachevar()
if timevar is None:
names = self.get_coord_names(var)
if names['tname'] is not None:
timevar = Timevar(self.nc, names["tname"])
else:
timevar = None
if use_cache is True:
self._coordcache[var].time = timevar
return timevar
def test_timevar_hfradar_days_values(self):
datafile = os.path.join(data_path, "marcooshfradar20120331.nc")
# Manually extract
# time:units = days since 2001-01-01 00:00:00
dt = datetime(2001, 1, 1, tzinfo=pytz.utc)
ds = netCDF4.Dataset(datafile)
data = ds.variables['time'][:]
# Timevar extract
tvar = Timevar(datafile, name='time', tzinfo=pytz.utc)
assert np.allclose(data, tvar.days)
# Now compare datetimes
data = data.tolist()
jds = []
for x in data:
jds.append(dt + timedelta(days=x))
assert (jds == tvar.dates).all()
ds.close()