if useEMG:
print('using EMG.')
label4EMG = params.label4withEMG
else:
label4EMG = params.label4withoutEMG
print('not using EMG')
print('pastStageLookUpNum = ' + str(pastStageLookUpNum))
#---------------
# read pickled files that stores EEG and stage labels
prefix = 'eegAndStage'
fileIDs = getFileIDs(eegDir, prefix)
### fileIDs = ['HET-NR-D0717', 'DBL-NO-D1473', 'HET-NO-D0905']
for fileID in fileIDs:
print('fileID = ' + str(fileID))
featureFileName = params.featureFilePrefix + '.' + params.extractorType + '.' + label4EMG + '.' + fileID + '.pkl'
if oFilter.isOutlier(fileID):
print('file ' + fileID + ' is an outlier, so skipping.')
else:
flag4extraction = 1
if option != '-o':
for fileName in listdir(featureDir):
if fileName == featureFileName:
flag4extraction = 0
print(featureFileName + ' already exists, so skipping')
break
from classifierTrainer import trainClassifier
from writeResults import writePredictions
orderMin = 0
orderMax = 8
# orderMax = 1
paramID = 0
params = ParameterSetup()
pickledDir = params.pickledDir
paramDir = params.pickledDir
classifierType = params.classifierType
classifierParams = params.classifierParams
eegFilePrefix = 'eegAndStage'
fileIDs = getFileIDs(pickledDir, eegFilePrefix)
print('classifier type = ' + str(classifierType))
print('useEMG = ' + str(params.useEMG))
print('emgTimeFrameNum = ' + str(params.emgTimeFrameNum))
print('binWidth4freqHisto = ' + str(params.binWidth4freqHisto))
stageLabels = params.stageLabels4evaluation
labelNum = len(stageLabels)
fileNum = len(fileIDs)
orderNum = orderMax - orderMin + 1
sensitivityMat = np.zeros((orderNum, labelNum), dtype=float)
specificityMat = np.zeros((orderNum, labelNum), dtype=float)
accuracyMat = np.zeros((orderNum, labelNum), dtype=float)
precisions = np.zeros((orderNum), dtype=float)
from parameterSetup import ParameterSetup
from fileManagement import getFileIDs
#---------------
# main
splitID = ''.join(
random.choice(string.ascii_uppercase + string.digits) for _ in range(4))
print('splitID =', splitID)
args = sys.argv
blockNum = int(args[1])
params = ParameterSetup()
# printMetadata(params)
prefix = 'eegAndStage'
fileIDs = getFileIDs(params.pickledDir, prefix)
recordNum = len(fileIDs)
print('recordNum =', recordNum)
# blockSize = np.int(np.floor(recordNum / blockNum))
blockSize = np.int(np.ceil(recordNum / blockNum))
print('blockSize =', blockSize)
print('fileIDs =', fileIDs[:5])
random.shuffle(fileIDs)
print('shuffled fileIDs =', fileIDs[:5])
blocks = []
for blockID in range(blockNum):
blocks.append(fileIDs[(blockID * blockSize):((blockID + 1) * blockSize)])
with open(params.pickledDir + '/blocks_of_records.' + splitID + '.csv',