def OneKfold(i=0, saveloclist=None):
modelloc = TrainModel(idfold=i, saveloclist=saveloclist)
PredictKFold(modelloc,
settings.options.dbfile,
settings.options.outdir,
kfolds=settings.options.kfolds,
idfold=settings.options.idfold)
def OneKfold(i=0, datadict=None):
k = settings.options.kfolds
modelloc = TrainModel(idfold=i)
(train_set, test_set, valid_set) = GetSetupKfolds(settings.options.dbfile,
k, i)
sumscore = 0
sumscorefloat = 0
for idtest in test_set:
baseloc = '%s/%03d/%03d' % (settings.options.outdir, k, i)
imgloc = '%s/%s' % (settings.options.rootlocation,
datadict[idtest]['image'])
segloc = '%s/%s' % (settings.options.rootlocation,
datadict[idtest]['label'])
outloc = '%s/label-%04d.nii.gz' % (baseloc, idtest)
if settings.options.numepochs > 0 and (
settings.options.makepredictions
or settings.options.makedropoutmap):
if settings.options.makepredictions:
predseg, predfloat = PredictModel(model=modelloc,
image=imgloc,
outdir=outloc)
else:
predseg, predfloat = PredictDropout(model=modelloc,
image=imgloc,
outdir=outloc,
seg=segloc)
def OneKfold(i=0, datadict=None):
k = settings.options.kfolds
modelloc = TrainModel(idfold=i)
(train_set,test_set) = GetSetupKfolds(settings.options.dbfile, k, i)
print('train set',train_set)
print('test set',test_set)
for idtest in test_set:
baseloc = '%s/%03d/%03d' % (settings.options.outdir, k, i)
imgloc = '%s/%s' % (settings.options.rootlocation, datadict[idtest]['image'])
segloc = '%s/%s' % (settings.options.rootlocation, datadict[idtest]['label'])
outloc = '%s/label-%04d.nii.gz' % (baseloc, idtest)
sumscore = 0
def OneKfold(i=0, datadict=None):
k = settings.options.kfolds
modelloc = TrainModel(idfold=i)
(train_set,test_set) = GetSetupKfolds(settings.options.dbfile, k, i)
print('train set',train_set)
print('test set',test_set)
sumscore = 0
sumscorefloat = 0
for idtest in test_set:
baseloc = '%s/%03d/%03d' % (settings.options.outdir, k, i)
imgloc = '%s/%s' % (settings.options.rootlocation, datadict[idtest]['image'])
segloc = '%s/%s' % (settings.options.rootlocation, datadict[idtest]['label'])
outloc = '%s/label-%04d.nii.gz' % (baseloc, idtest)
if settings.options.numepochs > 0 and (settings.options.makepredictions or settings.options.makedropoutmap) # as I train
imagepredict = nib.load(imgloc)
imageheader = imagepredict.header
numpypredict = imagepredict.get_data().astype(settings.IMG_DTYPE)
allseg = nib.load(segloc).get_data().astype(settings.SEG_DTYPE)
liver_idx = allseg > 0
tumor_idx = allseg > 1
seg_liver = np.zeros_like(allseg)
seg_liver[liver_idx] = 1
seg_tumor = np.zeros_like(allseg)
seg_tumor[tumor_idx] = 1
image_liver = seg_liver*numpypredict - (1.0 - seg_liver)
image_liver = image_liver.astype(settings.IMG_DTYPE)
if settings.options.makepredictions:
predseg, predfloat = PredictModelFromNumpy(model=modelloc, image=image_liver, imageheader=imageheader, outdir=outloc )
else:
predseg, predfloat = PredictDropoutFromNumpy(model=modelloc, image=image_liver, imageheader=imageheader, outdir=outloc)
score_float = dsc_l2_3D(seg_tumor.astype(settings.IMG_DTYPE), predfloat)
sumscorefloat += score_float
print(idtest, "\t", sumscorefloat)
print(k, " avg dice:\t", sumscorefloat/len(test_set))
def OneKfold(i=0, datadict=None):
k = settings.options.kfolds
modelloc = TrainModel(idfold=i)
(train_set, test_set) = GetSetupKfolds(settings.options.dbfile, k, i)
for idtest in test_set:
baseloc = '%s/%03d/%03d' % (settings.options.outdir, k, i)
imgloc = '%s/%s' % (settings.options.rootlocation,
datadict[idtest]['image'])
segloc = '%s/%s' % (settings.options.rootlocation,
datadict[idtest]['label'])
outloc = '%s/label-%04d.nii.gz' % (baseloc, idtest)
sumscore = 0
if settings.options.numepochs > 0 and settings.options.makepredictions: # doing K-fold prediction as I train
imagepredict = nib.load(imgloc)
imageheader = imagepredict.header
numpypredict = imagepredict.get_data().astype(settings.IMG_DTYPE)
allseg = nib.load(segloc).get_data().astype(settings.SEG_DTYPE)
liver_idx = allseg > 0
tumor_idx = allseg > 1
seg_liver = np.zeros_like(allseg)
seg_liver[liver_idx] = 1
seg_tumor = np.zeros_like(allseg)
seg_tumor[tumor_idx] = 1
image_liver = seg_liver * numpypredict - 100.0 * (1.0 - seg_liver)
image_liver = image_liver.astype(settings.IMG_DTYPE)
predseg = PredictModelFromNumpy(model=modelloc,
image=image_liver,
imageheader=imageheader,
outdir=outloc)
score = dsc_int_3D(seg_tumor, predseg)
sumscore += score
print(idtest, "\t", score)
print(k, " avg dice:\t", sumscore / len(test_set))
import settings
from settings import process_options, perform_setup
(options, args) = process_options()
IMG_DTYPE, SEG_DTYPE, _globalnpfile, _globalexpectedpixel, _nx, _ny = perform_setup(
options)
print('database file: %s ' % settings._globalnpfile)
from setupmodel import GetDataDictionary, BuildDB
from trainmodel import TrainModel
from predictmodel import PredictModel
from kfolds import OneKfold, Kfold
if options.builddb:
BuildDB()
if options.kfolds > 1:
if options.idfold > -1:
databaseinfo = GetDataDictionary(options.dbfile)
OneKfold(i=options.idfold, datadict=databaseinfo)
else:
Kfold()
if options.trainmodel and options.kfolds == 1: # no kfolds, i.e. k=1
TrainModel()
if options.predictmodel:
PredictModel()
if ((not options.builddb) and (not options.trainmodel)
and (not options.predictmodel) and (options.kfolds == 1)):
print("parser error")
start = time.time()
align = NaiveDlib(args.dlibFaceMean, args.dlibFacePredictor)
#net = openface.TorchWrap(args.networkModel, imgDim=args.imgDim, cuda=args.cuda)
if args.verbose:
print("Loading the dlib and OpenFace models took {} seconds.".format(
time.time() - start))
###################################################################################
from frameextractor import FrameExtractor
from labelframe import LabelFrame
from trainmodel import TrainModel
FE = FrameExtractor()
LF = LabelFrame()
TM = TrainModel()
print "====== Welcome to DriveSmart Training Module ======"
print "Enter the path to the video folder:"
videoPath = raw_input('--->')
print "Enter the text file containing list of videos with closed eyes:"
closedTextPath = raw_input('--->')
print "Enter the text file containing list of videos with open eyes:"
openTextPath = raw_input('--->')
#Extracting Frames
FE.setVideoFileName(videoPath)
FE.fetchFrame(closedTextPath, 0)
FE.fetchFrame(openTextPath, 1)
#Extracting Eye Patches
print('no list of tumor .npy files given for training')
quit()
else:
saveloclist = options.datafiles_tumor
else:
print('not specified liver vs tumor')
quit()
print('files already generated: using', saveloclist)
if options.kfolds > 1:
if options.idfold > -1:
OneKfold(i=options.idfold, saveloclist=saveloclist)
else:
Kfold(saveloclist=saveloclist)
else:
TrainModel(saveloclist=saveloclist)
if options.predictmodel:
if options.predictfromcsv:
PredictCSV(modelloc=options.predictmodel,
outdir=options.outdir,
indir=options.predictfromcsv)
else:
PredictNifti(model,
options.outdir + '/predictions/pred',
options.predictimage,
segloc=None)
PredictModel()
# To bypass certificate errors
try:
_create_unverified_https_context = ssl._create_unverified_context
except AttributeError:
pass
else:
ssl._create_default_https_context = _create_unverified_https_context
app = Flask(__name__)
# Class Objects
data_processor = DataProcess()
train_model = TrainModel()
# Pull all available data starting year 2000 until today
start_date = datetime(2000, 1, 1)
end_date = datetime.date(datetime.today())
# Input Arguments
apikey = 'JXYRIAIGOFQQEYU6'
apikey2 = 'OMX6JTTUJ7VZ4MOJ'
apikey3 = 'O4170I9AFRMU1MWM'
apikey4 = 'L7WQ5800OSKGRWRD'
symbol = 'SPY'
interval = 'daily'
interval_vwap = '60min'
time_period_cci_5 = '5'
time_period_cci_20 = '20'
