def _get_data_for_psd(self, inquiry_timing):
# get data from the DAQ
raw_data, triggers, target_info = process_data_for_decision(
inquiry_timing,
self.daq,
self.window,
self.parameters,
self.rsvp.first_stim_time,
self.static_offset,
buf_length=self.trial_length)
# filter it
notch_filterted_data = notch.notch_filter(raw_data, self.fs,
self.notch_filter_frequency)
filtered_data = bandpass.butter_bandpass_filter(
notch_filterted_data,
self.filter_low,
self.filter_high,
self.fs,
order=self.filter_order)
data = downsample.downsample(filtered_data,
factor=self.downsample_rate)
letters, times, target_info = self.letter_info(triggers, target_info)
# reshape with the filtered data with our desired window length
reshaped_data, _, _, _ = trial_reshaper(target_info,
times,
data,
fs=self.fs,
k=self.downsample_rate,
mode='calibration',
channel_map=self.channel_map,
trial_length=self.trial_length)
return reshaped_data
def test_trial_reshaper_calibration(self):
(reshaped_trials, labels,
num_of_seq, trials_per_sequence) = trial_reshaper(
trial_target_info=self.target_info,
timing_info=self.timing_info, eeg_data=self.eeg,
fs=256, k=2, mode='calibration', channel_map=self.channel_map)
self.assertTrue(
len(reshaped_trials) == self.channel_number,
f'len is {len(reshaped_trials)} not {self.channel_number}')
self.assertEqual(len(reshaped_trials[0]), 3)
self.assertEqual(num_of_seq, 1)
self.assertEqual(trials_per_sequence, 3)
def evaluate_sequence(self, raw_data, triggers, target_info,
window_length):
"""Once data is collected, infers meaning from the data.
Args:
raw_data(ndarray[float]): C x L eeg data where C is number of
channels and L is the signal length
triggers(list[tuple(str,float)]): triggers e.g. ('A', 1)
as letter and flash time for the letter
target_info(list[str]): target information about the stimuli
window_length(int): The length of the time between stimuli presentation
"""
letters, times, target_info = self.letter_info(triggers, target_info)
# Remove 60hz noise with a notch filter
notch_filter_data = notch.notch_filter(
raw_data,
self.sampling_rate,
frequency_to_remove=self.notch_filter_frequency)
# bandpass filter from 2-45hz
filtered_data = bandpass.butter_bandpass_filter(
notch_filter_data,
self.filter_low,
self.filter_high,
self.sampling_rate,
order=self.filter_order)
# downsample
data = downsample.downsample(filtered_data,
factor=self.downsample_rate)
x, _, _, _ = trial_reshaper(target_info,
times,
data,
fs=self.sampling_rate,
k=self.downsample_rate,
mode=self.mode,
channel_map=self.channel_map,
trial_length=window_length)
lik_r = inference(x, letters, self.signal_model, self.alp)
prob = self.conjugator.update_and_fuse({'ERP': lik_r})
decision, arg = self.decision_maker.decide(prob)
if 'stimuli' in arg:
sti = arg['stimuli']
else:
sti = None
return decision, sti
def offline_analysis(data_folder: str = None,
parameters: dict = {},
alert_finished: bool = True):
""" Gets calibration data and trains the model in an offline fashion.
pickle dumps the model into a .pkl folder
Args:
data_folder(str): folder of the data
save all information and load all from this folder
parameter(dict): parameters for running offline analysis
alert_finished(bool): whether or not to alert the user offline analysis complete
How it Works:
- reads data and information from a .csv calibration file
- reads trigger information from a .txt trigger file
- filters data
- reshapes and labels the data for the training procedure
- fits the model to the data
- uses cross validation to select parameters
- based on the parameters, trains system using all the data
- pickle dumps model into .pkl file
- generates and saves offline analysis screen
- [optional] alert the user finished processing
"""
if not data_folder:
data_folder = load_experimental_data()
mode = 'calibration'
trial_length = parameters.get('collection_window_after_trial_length')
raw_dat, _, channels, type_amp, fs = read_data_csv(
data_folder + '/' + parameters.get('raw_data_name', 'raw_data.csv'))
log.info(f'Channels read from csv: {channels}')
log.info(f'Device type: {type_amp}')
downsample_rate = parameters.get('down_sampling_rate', 2)
# Remove 60hz noise with a notch filter
notch_filter_data = notch.notch_filter(raw_dat, fs, frequency_to_remove=60)
# bandpass filter from 2-45hz
filtered_data = bandpass.butter_bandpass_filter(notch_filter_data,
2,
45,
fs,
order=2)
# downsample
data = downsample.downsample(filtered_data, factor=downsample_rate)
# Process triggers.txt
triggers_file = parameters.get('trigger_file_name', 'triggers.txt')
_, t_t_i, t_i, offset = trigger_decoder(
mode=mode, trigger_path=f'{data_folder}/{triggers_file}')
static_offset = parameters.get('static_trigger_offset', 0)
offset = offset + static_offset
# Channel map can be checked from raw_data.csv file.
# read_data_csv already removes the timespamp column.
channel_map = analysis_channels(channels, type_amp)
x, y, _, _ = trial_reshaper(t_t_i,
t_i,
data,
mode=mode,
fs=fs,
k=downsample_rate,
offset=offset,
channel_map=channel_map,
trial_length=trial_length)
k_folds = parameters.get('k_folds', 10)
model, auc = train_pca_rda_kde_model(x, y, k_folds=k_folds)
log.info('Saving offline analysis plots!')
# After obtaining the model get the transformed data for plotting purposes
model.transform(x)
generate_offline_analysis_screen(
x,
y,
model=model,
folder=data_folder,
down_sample_rate=downsample_rate,
fs=fs,
save_figure=True,
show_figure=False,
channel_names=analysis_channel_names_by_pos(channels, channel_map))
log.info('Saving the model!')
with open(data_folder + f'/model_{auc}.pkl', 'wb') as output:
pickle.dump(model, output)
if alert_finished:
offline_analysis_tone = parameters.get('offline_analysis_tone')
play_sound(offline_analysis_tone)
return model
