def getNonMonitoredTarget(webpageTrainSource):
if config.BUILD_ALL_WEBPAGES:
webpageTrain = Datastore.getWebpagesWangTor([101], 1, 2)
webpageTrain = webpageTrain[0]
config.ALL_WEBPAGES.append(webpageTrain)
config.BUILD_ALL_WEBPAGES = False
config.LARGEST_WEBPAGE = webpageTrain
#print "src id: " + str(webpageTrainSource.getId())
#print "largest target id: " + str(config.LARGEST_WEBPAGE.getId())
#print "---"
return config.LARGEST_WEBPAGE
elif config.DATA_SOURCE == 1 or config.DATA_SOURCE == 2:
webpageTrain = Datastore.getWebpagesHerrmann(
[webpageId], seed - config.NUM_TRAINING_TRACES,
seed)
webpageTest = Datastore.getWebpagesHerrmann(
[webpageId], seed,
seed + config.NUM_TESTING_TRACES)
elif config.DATA_SOURCE == 5:
if not unMonitoredWebpageIdsObj.__contains__(
webpageId):
# this block of code can be applied to either a closed or open-world
if config.COVARIATE_SHIFT == 0: # Normal case
# monitored webpage so we take instances for training and testing as we do regularly
webpageTrain = Datastore.getWebpagesWangTor(
[webpageId],
seed - config.NUM_TRAINING_TRACES, seed)
webpageTest = Datastore.getWebpagesWangTor(
[webpageId], seed,
seed + config.NUM_TESTING_TRACES)
else:
# span time training/testing
# monitored webpage so we take instances for training and testing as we do regularly
webpageTrain = Datastore.getWebpagesWangTor(
[webpageId], 0, config.NUM_TRAINING_TRACES)
webpageTest = Datastore.getWebpagesWangTor(
[webpageId], config.NUM_TRAINING_TRACES +
config.COVARIATE_SHIFT,
config.NUM_TRAINING_TRACES +
config.COVARIATE_SHIFT +
config.NUM_TESTING_TRACES)
elif config.DATA_SOURCE == 1 or config.DATA_SOURCE == 2:
webpageTrain = Datastore.getWebpagesHerrmann(
[webpageId], seed - config.NUM_TRAINING_TRACES,
seed)
webpageTest = Datastore.getWebpagesHerrmann(
[webpageId], seed,
seed + config.NUM_TESTING_TRACES)
elif config.DATA_SOURCE == 5:
if not unMonitoredWebpageIdsObj.__contains__(
webpageId):
# this block of code can be applied to either a closed or open-world
if config.COVARIATE_SHIFT == 0: # Normal case
# monitored webpage so we take instances for training and testing as we do regularly
webpageTrain = Datastore.getWebpagesWangTor(
[webpageId],
seed - config.NUM_TRAINING_TRACES,
seed)
webpageTest = Datastore.getWebpagesWangTor(
[webpageId], seed,
seed + config.NUM_TESTING_TRACES)
else:
# span time training/testing
# monitored webpage so we take instances for training and testing as we do regularly
#webpageTrain = Datastore.getWebpagesWangTor( [webpageId], 0, config.NUM_TRAINING_TRACES )
#webpageTest = Datastore.getWebpagesWangTor( [webpageId], config.NUM_TRAINING_TRACES+config.COVARIATE_SHIFT, config.NUM_TRAINING_TRACES+config.COVARIATE_SHIFT+config.NUM_TESTING_TRACES )
endSlideTrain = config.START_SLIDE_TRAIN + config.NUM_TRAINING_TRACES
webpageTrain = Datastore.getWebpagesWangTor(
[webpageId], config.START_SLIDE_TRAIN,
endSlideTrain)
webpageTest = Datastore.getWebpagesWangTor(
