Feature, EEG_Recording = FeatureExtraction(Sample, Feature,
EEG_Recording)
# Signal Start
FIFO.Rewrite(Trigger_Log, "Trial Start")
print 'Write "Trial Start"'
Trigger = np.concatenate((Trigger, np.array([[len(Feature),
TRIAL_START]])))
Max_Power = np.percentile(
Feature[range(
len(Feature) - 5 * TaskSetting['Fixation'] -
5 * TaskSetting['Trial Start'], len(Feature))], 90)
print Max_Power
if not Max_Power == Min_Power:
Classifier, Offset = Process.Linear_Regression(
(Min_Power, Max_Power), (Movement_Range_Min, Movement_Range_Max))
print Classifier
else:
print 'ERROR'
Classifier = 0
Offset = 0
# Clean up the Queue to keep Data in real time
if Data_Queue.empty():
print 'No Left Behind'
else:
while not Data_Queue.empty():
Sample = Data_Queue.get()
Feature, EEG_Recording = FeatureExtraction(Sample, Feature,
EEG_Recording)
Current_Index = len(Feature)
screenWidth, screenHeight = gui.size()
gui.FAILSAFE = False
MovementRange_X = [-5.0 * (screenWidth / 100.0), 5.0 * (screenWidth / 100.0)]
MovementRange_Y = [-5.0 * (screenHeight / 100.0), 5.0 * (screenHeight / 100.0)]
while True:
if not Data_Queue.empty():
Sample = Data_Queue.get()
Sample[:, range(1, 9)] = Process.Common_Average_Reference(
Sample[:, range(1, 9)])
Feature1.append(
Process.PowerExtraction(Sample[:, 4], FS, Frequency_Range))
Feature2.append(
Process.PowerExtraction(Sample[:, 7], FS, Frequency_Range))
Classifier_X, Offset_X = Process.Linear_Regression(
Feature1, MovementRange_X)
Classifier_Y, Offset_Y = Process.Linear_Regression(
Feature2, MovementRange_Y)
if not Classifier_X == np.inf and not Classifier_Y == np.inf:
break
while len(Feature1) < 70:
if not Data_Queue.empty():
Sample = Data_Queue.get()
Sample[:, range(1, 9)] = Process.Common_Average_Reference(
Sample[:, range(1, 9)])
Feature1.append(
Process.PowerExtraction(Sample[:, 4], FS, Frequency_Range))
Feature2.append(
Process.PowerExtraction(Sample[:, 7], FS, Frequency_Range))
Distance_X = Feature1[len(Feature1) - 1] * Classifier_X + Offset_X
