def _stop_brainflow(self):
"""This functions kills the brainflow backend and saves the data to a CSV file."""
# Collect session data and kill session
data = self.board.get_board_data() # will clear board buffer
self.board.stop_stream()
self.board.release_session()
# transform data for saving
data = data.T # transpose data
ch_names = BRAINFLOW_CHANNELS[self.device_name]
num_channels = len(ch_names)
eeg_data = data[:, 1:num_channels + 1]
timestamps = data[:, -1]
# Create a column for the stimuli to append to the EEG data
stim_array = create_stim_array(timestamps, self.markers)
timestamps = timestamps[
..., None] # Add an additional dimension so that shapes match
total_data = np.append(timestamps, eeg_data, 1)
total_data = np.append(total_data, stim_array,
1) # Append the stim array to data.
# Subtract five seconds of settling time from beginning
total_data = total_data[5 * self.sfreq:]
data_df = pd.DataFrame(total_data,
columns=['timestamps'] + ch_names + ['stim'])
data_df.to_csv(self.save_fn, index=False)
def _stop_brainflow(self):
"""This functions kills the brainflow backend and saves the data to a CSV file."""
# Collect session data and kill session
data = self.board.get_board_data() # will clear board buffer
self.board.stop_stream()
self.board.release_session()
# transform data for saving
data = data.T # transpose data
# get the channel names for EEG data
if self.brainflow_id == BoardIds.GANGLION_BOARD.value:
# if a ganglion is used, use recommended default EEG channel names
ch_names = ['fp1', 'fp2', 'tp7', 'tp8']
else:
# otherwise select eeg channel names via brainflow API
ch_names = BoardShim.get_eeg_names(self.brainflow_id)
# pull EEG channel data via brainflow API
eeg_data = data[:, BoardShim.get_eeg_channels(self.brainflow_id)]
timestamps = data[:, BoardShim.get_timestamp_channel(self.brainflow_id)]
# Create a column for the stimuli to append to the EEG data
stim_array = create_stim_array(timestamps, self.markers)
timestamps = timestamps[..., None] # Add an additional dimension so that shapes match
total_data = np.append(timestamps, eeg_data, 1)
total_data = np.append(total_data, stim_array, 1) # Append the stim array to data.
# Subtract five seconds of settling time from beginning
total_data = total_data[5 * self.sfreq:]
data_df = pd.DataFrame(total_data, columns=['timestamps'] + ch_names + ['stim'])
data_df.to_csv(self.save_fn, index=False)
def get_data(self) -> pd.DataFrame:
from eegnb.devices.utils import create_stim_array
data = self.board.get_board_data() # will clear board buffer
# transform data for saving
data = data.T # transpose data
print(data)
# get the channel names for EEG data
if self.brainflow_id == BoardIds.GANGLION_BOARD.value:
# if a ganglion is used, use recommended default EEG channel names
ch_names = ["fp1", "fp2", "tp7", "tp8"]
else:
# otherwise select eeg channel names via brainflow API
ch_names = BoardShim.get_eeg_names(self.brainflow_id)
# pull EEG channel data via brainflow API
eeg_data = data[:, BoardShim.get_eeg_channels(self.brainflow_id)]
timestamps = data[:,
BoardShim.get_timestamp_channel(self.brainflow_id)]
# Create a column for the stimuli to append to the EEG data
stim_array = create_stim_array(timestamps, self.markers)
timestamps = timestamps[
..., None] # Add an additional dimension so that shapes match
total_data = np.append(timestamps, eeg_data, 1)
total_data = np.append(total_data, stim_array,
1) # Append the stim array to data.
# Subtract five seconds of settling time from beginning
# total_data = total_data[5 * self.sfreq :]
df = pd.DataFrame(total_data,
columns=["timestamps"] + ch_names + ["stim"])
return df
def _stop_brainflow(self):
"""This functions kills the brainflow backend and saves the data to a CSV file."""
# Collect session data and kill session
data = self.board.get_board_data() # will clear board buffer
self.board.stop_stream()
self.board.release_session()
# Extract relevant metadata from board
ch_names, eeg_data, timestamps = self._brainflow_extract(data)
# Create a column for the stimuli to append to the EEG data
stim_array = create_stim_array(timestamps, self.markers)
timestamps = timestamps[..., None]
# Add an additional dimension so that shapes match
total_data = np.append(timestamps, eeg_data, 1)
# Append the stim array to data.
total_data = np.append(total_data, stim_array, 1)
# Subtract five seconds of settling time from beginning
total_data = total_data[5 * self.sfreq:]
data_df = pd.DataFrame(total_data,
columns=["timestamps"] + ch_names + ["stim"])
data_df.to_csv(self.save_fn, index=False)