# EPOCH EXTRACTION CONFIG:
EVENT_IDS = [1, 2]
T_MIN, T_MAX = 2.5, 4.5 # time before event, time after event
CSP_N = 12
ap = Approach()
ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF, FILT_ORDER, CSP_N,
EVENT_IDS, T_MIN, T_MAX)
ap.setPathToCal(DATA_CAL_PATH, CAL_EVENTS_PATH)
ap.setPathToVal(DATA_VAL_PATH, VAL_EVENTS_PATH)
ap.setValidChannels([-1])
ap.set_balance_epochs(False)
autoscore = ap.trainModel()
valscore = ap.validateModel()
print autoscore
print valscore
## test on single epoch
# i = 0
# while i < len(ap.labels_cal):
# epoch_number = i
# EPOCH EXTRACTION CONFIG: EVENT_IDS = [1,2] T_MIN, T_MAX = 2.5,4.5 # time before event, time after event CSP_N = 12 ap = Approach() ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF, FILT_ORDER, CSP_N, EVENT_IDS, T_MIN, T_MAX) ap.setPathToCal(DATA_CAL_PATH, CAL_EVENTS_PATH) ap.setPathToVal(DATA_VAL_PATH, VAL_EVENTS_PATH) ap.setValidChannels([-1]) ap.set_balance_epochs(False) autoscore = ap.trainModel() valscore = ap.validateModel() print autoscore print valscore ## test on single epoch # i = 0 # while i < len(ap.labels_cal): # epoch_number = i
LOWER_CUTOFF = 8.
UPPER_CUTOFF = 30.
FILT_ORDER = 7
# EPOCH EXTRACTION CONFIG:
EVENT_IDS = [769, 770, 771, 772]
T_MIN, T_MAX = 0.5, 2.5 # time before event, time after event
CSP_N = 8
ap = Approach()
ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF, FILT_ORDER, CSP_N,
EVENT_IDS, T_MIN, T_MAX)
ap.setValidChannels(range(22))
data, ev = ap.loadData(DATA_PATH, EVENTS_PATH)
epochs, labels = ap.loadEpochs(data, ev)
epochs = ap.preProcess(epochs)
idx_left = np.where(labels == EVENT_IDS[0])[0]
idx_right = np.where(labels == EVENT_IDS[1])[0]
idx_foot = np.where(labels == EVENT_IDS[2])[0]
idx_tongue = np.where(labels == EVENT_IDS[3])[0]
new_data = np.zeros([1, epochs.shape[1]])
new_events = np.zeros([1, 2])
T_MIN += increment
T_MAX += increment
t.extend([T_MIN])
CSP_N = 12
ap = Approach()
ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF, FILT_ORDER,
CSP_N, EVENT_IDS, T_MIN, T_MAX)
ap.setPathToCal(DATA_CAL_PATH, CAL_EVENTS_PATH)
ap.setValidChannels(range(16))
ap.define_bad_epochs(50, None)
data, events = ap.loadData(DATA_CAL_PATH, CAL_EVENTS_PATH)
ref_channel = 8 # fcz
data = ap.preProcess(data)
data = data[:, :] - data[ref_channel]
# nch = data.shape[0]
# Id = np.identity(nch)
# W = Id - (1.0 / nch) * np.dot(Id, Id.T)
# data = np.dot(W, data)
LOWER_CUTOFF = 8.
UPPER_CUTOFF = 30.
FILT_ORDER = 7
# EPOCH EXTRACTION CONFIG:
EVENT_IDS = [769, 770, 771, 772]
T_MIN, T_MAX = 0.5, 2.5 # time before event, time after event
CSP_N = 8
ap = Approach()
ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF,
FILT_ORDER, CSP_N, EVENT_IDS, T_MIN, T_MAX)
ap.setValidChannels(range(22))
data, ev = ap.loadData(DATA_PATH, EVENTS_PATH)
epochs, labels = ap.loadEpochs(data, ev)
epochs = ap.preProcess(epochs)
idx_left = np.where(labels == EVENT_IDS[0])[0]
idx_right = np.where(labels == EVENT_IDS[1])[0]
idx_foot = np.where(labels == EVENT_IDS[2])[0]
idx_tongue = np.where(labels == EVENT_IDS[3])[0]
new_data = np.zeros([1, epochs.shape[1]])
while T_MAX < end_t: T_MIN += increment T_MAX += increment t.extend([T_MIN]) CSP_N = 12 ap = Approach() ap.defineApproach(SAMPLING_FREQ, LOWER_CUTOFF, UPPER_CUTOFF, FILT_ORDER, CSP_N, EVENT_IDS, T_MIN, T_MAX) ap.setPathToCal(DATA_CAL_PATH, CAL_EVENTS_PATH) ap.setValidChannels(range(16)) ap.define_bad_epochs(50, None) data, events = ap.loadData(DATA_CAL_PATH, CAL_EVENTS_PATH) ref_channel = 8 # fcz data = ap.preProcess(data) data = data[:,:] - data[ref_channel] # nch = data.shape[0] # Id = np.identity(nch) # W = Id - (1.0 / nch) * np.dot(Id, Id.T) # data = np.dot(W, data)