def build_stimulus_table(stimulus_pkl_path, sync_h5_path, frame_time_strategy,
minimum_spontaneous_activity_duration,
extract_const_params_from_repr, drop_const_params,
maximum_expected_spontanous_activity_duration,
stimulus_name_map, column_name_map,
output_stimulus_table_path, output_frame_times_path,
fail_on_negative_duration, **kwargs):
stim_file = CamStimOnePickleStimFile.factory(stimulus_pkl_path)
sync_dataset = EcephysSyncDataset.factory(sync_h5_path)
frame_times = sync_dataset.extract_frame_times(
strategy=frame_time_strategy)
def seconds_to_frames(seconds):
return \
(np.array(seconds) + stim_file.pre_blank_sec) * \
stim_file.frames_per_second
minimum_spontaneous_activity_duration = (
minimum_spontaneous_activity_duration / stim_file.frames_per_second)
stimulus_tabler = functools.partial(
ephys_pre_spikes.build_stimuluswise_table,
seconds_to_frames=seconds_to_frames,
extract_const_params_from_repr=extract_const_params_from_repr,
drop_const_params=drop_const_params,
)
spon_tabler = functools.partial(
ephys_pre_spikes.make_spontaneous_activity_tables,
duration_threshold=minimum_spontaneous_activity_duration,
)
stim_table_full = ephys_pre_spikes.create_stim_table(
stim_file.stimuli, stimulus_tabler, spon_tabler)
stim_table_full = ephys_pre_spikes.apply_frame_times(
stim_table_full, frame_times, stim_file.frames_per_second, True)
output_validation.validate_epoch_durations(
stim_table_full, fail_on_negative_durations=fail_on_negative_duration)
output_validation.validate_max_spontaneous_epoch_duration(
stim_table_full, maximum_expected_spontanous_activity_duration)
stim_table_full = naming_utilities.collapse_columns(stim_table_full)
stim_table_full = naming_utilities.drop_empty_columns(stim_table_full)
stim_table_full = naming_utilities.standardize_movie_numbers(
stim_table_full)
stim_table_full = naming_utilities.add_number_to_shuffled_movie(
stim_table_full)
stim_table_full = naming_utilities.map_stimulus_names(
stim_table_full, stimulus_name_map)
stim_table_full = naming_utilities.map_column_names(
stim_table_full, column_name_map)
stim_table_full.to_csv(output_stimulus_table_path, index=False)
np.save(output_frame_times_path, frame_times, allow_pickle=False)
return {
"output_path": output_stimulus_table_path,
"output_frame_times_path": output_frame_times_path,
}
def build_opto_table(args):
opto_file = pd.read_pickle(args['opto_pickle_path'])
sync_file = EcephysSyncDataset.factory(args['sync_h5_path'])
start_times = sync_file.extract_led_times()
conditions = [str(item) for item in opto_file['opto_conditions']]
levels = opto_file['opto_levels']
assert len(conditions) == len(levels)
if len(start_times) > len(conditions):
raise ValueError(
f"there are {len(start_times) - len(conditions)} extra optotagging sync times!"
)
optotagging_table = pd.DataFrame({
'start_time': start_times,
'condition': conditions,
'level': levels
})
optotagging_table = optotagging_table.sort_values(by='start_time', axis=0)
stop_times = []
names = []
conditions = []
for ii, row in optotagging_table.iterrows():
condition = args["conditions"][row["condition"]]
stop_times.append(row["start_time"] + condition["duration"])
names.append(condition["name"])
conditions.append(condition["condition"])
optotagging_table["stop_time"] = stop_times
optotagging_table["stimulus_name"] = names
optotagging_table["condition"] = conditions
optotagging_table["duration"] = optotagging_table[
"stop_time"] - optotagging_table["start_time"]
optotagging_table.to_csv(args['output_opto_table_path'], index=False)
return {'output_opto_table_path': args['output_opto_table_path']}
def test_factory():
with mock.patch('allensdk.brain_observatory.sync_dataset.Dataset.load') as p:
dataset = EcephysSyncDataset.factory('foo')
p.assert_called_with('foo')