def load_session_and_get_fr_df(self):
session = EcephysSession.from_nwb_path(self.nwb_path,
api_kwargs={
"amplitude_cutoff_maximum":
0.1,
"presence_ratio_minimum":
0.9,
"isi_violations_maximum":
0.5
})
self.units_df = session.units
self.stim_table = session.get_stimulus_table(self.stim_name)
stim_pres_ids = self.stim_table.index.values
tmp_binned_spt = session.presentationwise_spike_counts\
(bin_edges = self.bin_edges, stimulus_presentation_ids=stim_pres_ids, unit_ids = session.units.index.values)
num_pres, num_bins, num_cells = tmp_binned_spt.shape
print(num_pres, num_bins, num_cells)
tot_arr_fr_all = np.reshape(tmp_binned_spt.values,
(num_pres * num_bins, num_cells))
self.tot_fr_df = pd.DataFrame(tot_arr_fr_all.T,
index=session.units.index.values)
del tmp_binned_spt, session
def test_spikes_nwb1(nwb_path):
"""
This test was based on the file /allen/aibs/mat/ecephys_data/mouse412792.spikes.nwb. To run this test please copy
or create a link to it in this directory.
"""
if not nwb_path.exists():
pytest.skip()
# TODO: Convert this NWB 1 file into a NWB 2 file that way we can check that NWB Adaptors return the same data
# and computations (minus a few exceptions for missing NWB 1 data).
session = EcephysSession.from_nwb_path(path=str(nwb_path), nwb_version=1)
assert (isinstance(session.units, pd.DataFrame))
assert (len(session.units) == 1363)
print(session.stimulus_names)
assert (isinstance(session.stimulus_presentations, pd.DataFrame))
assert (len(session.stimulus_presentations) == 70390)
assert (len(session.get_stimulus_table(['Natural Images_5'])) == 5950)
assert (len(session.get_stimulus_table(['drifting_gratings_2'])) == 630)
assert (len(session.get_stimulus_table(['flash_250ms_1'])) == 150)
assert (len(session.get_stimulus_table(['gabor_20_deg_250ms_0'])) == 3645)
assert (len(session.get_stimulus_table(['natural_movie_one_three'
])) == 18000)
assert (len(session.get_stimulus_table(['natural_movie_three_four'
])) == 36000)
assert (len(session.get_stimulus_table(['spontaneous'])) == 15)
assert (len(session.get_stimulus_table(['static_gratings_6'])) == 6000)
assert (session.running_speed.shape[0] == 365700)
assert (len(session.spike_times.keys()) == 1363)
assert (len(session.mean_waveforms.keys()) == 1363)
one_waveform = next(iter(session.mean_waveforms.values()))
assert (isinstance(one_waveform, xr.DataArray))
assert (len(session.probes) == 6)
assert (len(session.channels) == 737)
pst = session.presentationwise_spike_times()
assert (isinstance(pst, pd.DataFrame) and len(pst) > 0)
cpc = session.conditionwise_spike_statistics(
stimulus_presentation_ids=session.stimulus_presentations.index.
values[:40])
assert (isinstance(cpc, pd.DataFrame) and len(cpc) > 0)
# %%
rfs = []
all_rf_unit_ids = []
for session_id, row in sessions.iterrows():
print(session_id)
nwb_path = cache_dir + '/session_' + str(session_id) + '/session_' + str(
session_id) + '.nwb'
session = EcephysSession.from_nwb_path(nwb_path,
api_kwargs={
"amplitude_cutoff_maximum":
np.inf,
"presence_ratio_minimum":
-np.inf,
"isi_violations_maximum":
np.inf,
"filter_by_validity": True
})
print(' Calculating RFs...')
RFs = ReceptiveFieldMapping(session).receptive_fields
rfs.append(RFs['spike_counts'].data)
all_rf_unit_ids.append(RFs.unit_id.values)
# %%
all_RFs = np.concatenate(rfs, axis=2)