def __init__(self,
min_num_seq,
max_num_seq,
signal_model=None,
fs=300,
k=2,
alp=None,
evidence_names=['LM', 'ERP'],
task_list=[('I_LOVE_COOKIES', 'I_LOVE_')],
lmodel=None,
is_txt_stim=True,
device_name='LSL',
device_channels=None,
stimuli_timing=[1, .2],
decision_threshold=0.8,
backspace_prob=0.05,
backspace_always_shown=False,
filter_high=45,
filter_low=2,
filter_order=2,
notch_filter_frequency=60):
self.conjugator = EvidenceFusion(evidence_names, len_dist=len(alp))
seq_constants = []
if backspace_always_shown and BACKSPACE_CHAR in alp:
seq_constants.append(BACKSPACE_CHAR)
self.decision_maker = DecisionMaker(
min_num_seq,
max_num_seq,
decision_threshold=decision_threshold,
state=task_list[0][1],
alphabet=alp,
is_txt_stim=is_txt_stim,
stimuli_timing=stimuli_timing,
seq_constants=seq_constants)
self.alp = alp
# non-letter target labels include the fixation cross and calibration.
self.nonletters = ['+', 'PLUS', 'calibration_trigger']
self.valid_targets = set(self.alp)
self.signal_model = signal_model
self.sampling_rate = fs
self.downsample_rate = k
self.filter_high = filter_high
self.filter_low = filter_low
self.filter_order = filter_order
self.notch_filter_frequency = notch_filter_frequency
self.mode = 'copy_phrase'
self.task_list = task_list
self.lmodel = lmodel
self.channel_map = analysis_channels(device_channels, device_name)
self.backspace_prob = backspace_prob
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
def __init__(self,
min_num_inq,
max_num_inq,
signal_model=None,
fs=300,
k=2,
alp=None,
evidence_names=['LM', 'ERP'],
task_list=[('I_LOVE_COOKIES', 'I_LOVE_')],
lmodel=None,
is_txt_stim=True,
device_name='LSL',
device_channels=None,
stimuli_timing=[1, .2],
decision_threshold=0.8,
backspace_prob=0.05,
backspace_always_shown=False,
filter_high=45,
filter_low=2,
filter_order=2,
notch_filter_frequency=60):
self.conjugator = EvidenceFusion(evidence_names, len_dist=len(alp))
inq_constants = []
if backspace_always_shown and BACKSPACE_CHAR in alp:
inq_constants.append(BACKSPACE_CHAR)
# Stimuli Selection Module
stopping_criteria = CriteriaEvaluator(
continue_criteria=[MinIterationsCriteria(min_num_inq)],
commit_criteria=[
MaxIterationsCriteria(max_num_inq),
ProbThresholdCriteria(decision_threshold)
])
# TODO: Parametrize len_query in the future releases!
stimuli_agent = NBestStimuliAgent(alphabet=alp, len_query=10)
self.decision_maker = DecisionMaker(
stimuli_agent=stimuli_agent,
stopping_evaluator=stopping_criteria,
state=task_list[0][1],
alphabet=alp,
is_txt_stim=is_txt_stim,
stimuli_timing=stimuli_timing,
inq_constants=inq_constants)
self.alp = alp
# non-letter target labels include the fixation cross and calibration.
self.nonletters = ['+', 'PLUS', 'calibration_trigger']
self.valid_targets = set(self.alp)
self.signal_model = signal_model
self.sampling_rate = fs
self.downsample_rate = k
self.filter_high = filter_high
self.filter_low = filter_low
self.filter_order = filter_order
self.notch_filter_frequency = notch_filter_frequency
self.mode = 'copy_phrase'
self.task_list = task_list
self.lmodel = lmodel
self.channel_map = analysis_channels(device_channels, device_name)
self.backspace_prob = backspace_prob
def __init__(self, win, daq, parameters, file_save):
super(RSVPInterInquiryFeedbackCalibration, self).__init__()
self._task = RSVPCalibrationTask(win, daq, parameters, file_save)
self.daq = daq
self.fs = self.daq.device_info.fs
self.alp = self._task.alp
self.rsvp = self._task.rsvp
self.parameters = parameters
self.file_save = file_save
self.enable_breaks = self._task.enable_breaks
self.window = self._task.window
self.stim_number = self._task.stim_number
self.stim_length = self._task.stim_length
self.is_txt_stim = self.rsvp.is_txt_stim
self.stimuli_height = self._task.stimuli_height
self.color = self._task.color
self.timing = self._task.timing
self.wait_screen_message = self._task.wait_screen_message
self.wait_screen_message_color = self._task.wait_screen_message_color
self.visual_feedback = LevelFeedback(display=self.window,
parameters=self.parameters,
clock=self._task.experiment_clock)
self.static_offset = self.parameters['static_trigger_offset']
self.nonletters = ['+', 'PLUS', 'calibration_trigger']
self.valid_targets = set(self.alp)
self.time_flash = self.parameters['time_flash']
self.downsample_rate = self.parameters['down_sampling_rate']
self.filtered_sampling_rate = self.fs / self.downsample_rate
self.device_name = self.daq.device_info.name
self.channel_map = analysis_channels(self.daq.device_info.channels,
self.device_name)
# EDIT ME FOR FEEDBACK CONFIGURATION
self.feedback_buffer_time = self.parameters['feedback_buffer_time']
self.feedback_line_color = self.parameters['feedback_line_color']
self.psd_method = PSD_TYPE.WELCH
# The channel used to calculate the PSD from RSVP inquiry.
self.psd_channel_index = self.PSD_CHANNEL_INDEX
# filter parameters
self.filter_low = self.parameters['filter_low']
self.filter_high = self.parameters['filter_high']
self.filter_order = self.parameters['filter_order']
self.notch_filter_frequency = self.parameters['notch_filter_frequency']
# get the feedback band of interest
self.psd_lower_limit = self.parameters['feedback_band_lower_limit']
self.psd_upper_limit = self.parameters['feedback_band_upper_limit']
# psd band of interest to use for feeback (low, high)
self.psd_export_band = (self.psd_lower_limit, self.psd_upper_limit)
# length of time to use for PSD calculation
self.trial_length = self.time_flash * self.stim_length
self.lvl_5_threshold = self.parameters['feedback_level_5_threshold']
self.lvl_4_threshold = self.parameters['feedback_level_4_threshold']
self.lvl_3_threshold = self.parameters['feedback_level_3_threshold']
self.lvl_2_threshold = self.parameters['feedback_level_2_threshold']
# true/false order is desceding from 5 -> 1 for level
self.feedback_descending = self.parameters['feedback_level_descending']