def prediction():
"""Returns predictions for a given dataset."""
# Missing arguments.
if len(sys.argv) < 4:
result = dict()
result['runid'] = str(int(time.time()))
result['status'] = estimator.GENERAL_ERROR
result['info'] = ['Missing arguments, you should set:\
- The model unique identifier\
- The directory to store all generated outputs\
- The file with samples to predict\
Received: ' + ' '.join(sys.argv)]
print(json.dumps(result))
sys.exit(result['status'])
modelid = sys.argv[1]
directory = sys.argv[2]
regressor = estimator.Regressor(modelid, directory)
result = regressor.predict_dataset(sys.argv[3])
print(json.dumps(result))
sys.exit(result['status'])
def training():
"""Trains a ML regressor."""
# Missing arguments.
if len(sys.argv) < 4:
result = dict()
result['runid'] = str(int(time.time()))
result['status'] = estimator.GENERAL_ERROR
result['info'] = [
'Missing arguments, you should set:\
- The model unique identifier\
- The directory to store all generated outputs\
- The training file\
Received: ' + ' '.join(sys.argv)
]
print(json.dumps(result))
sys.exit(result['status'])
modelid = sys.argv[1]
directory = sys.argv[2]
regressor = estimator.Regressor(modelid, directory)
result = regressor.train_dataset(sys.argv[3])
print(json.dumps(result))
sys.exit(result['status'])
def import():
"""Imports a trained classifier or regressor."""
modelid = sys.argv[1]
directory = sys.argv[2]
type = sys.argv[4]
if type=="classifier" :
binary_classifier = estimator.Binary(modelid, directory)
binary_classifier.import_classifier(sys.argv[3])
# An exception will be thrown before if it can be imported.
sys.exit(0)
else:
regressor = estimator.Regressor(modelid, directory)
regressor.import_classifier(sys.argv[3])
# An exception will be thrown before if it can be imported.
sys.exit(0)
def export():
"""Exports the classifier or regressor."""
modelid = sys.argv[1]
directory = sys.argv[2]
type = sys.argv[4]
if type=="classifier" :
binary_classifier = estimator.Binary(modelid, directory)
exportdir = binary_classifier.export_classifier(sys.argv[3])
if exportdir:
print(exportdir)
sys.exit(0)
sys.exit(1)
else :
regressor = estimator.Regressor(modelid, directory)
exportdir = regressor.export_regressor(sys.argv[3])
if exportdir:
print(exportdir)
sys.exit(0)
sys.exit(1)
def evaluation():
"""Evaluates the provided dataset."""
# Missing arguments.
if len(sys.argv) < 7:
result = dict()
result['runid'] = str(int(time.time()))
result['status'] = estimator.GENERAL_ERROR
result['info'] = [
'Missing arguments, you should set:\
- The model unique identifier\
- The directory to store all generated outputs\
- The training file\
- The msximum deviation\
- The number of times the evaluation will run (defaults to 100)\
Received: ' + ' '.join(sys.argv)
]
print(json.dumps(result))
sys.exit(result['status'])
modelid = sys.argv[1]
directory = sys.argv[2]
regressor = estimator.Regressor(modelid, directory)
if len(sys.argv) > 7:
trained_model_dir = sys.argv[7]
else:
trained_model_dir = False
result = regressor.evaluate_dataset(sys.argv[3], float(sys.argv[4]),
int(sys.argv[5]), trained_model_dir)
print(json.dumps(result))
sys.exit(result['status'])