def train_by_bucket_data(*args, **kwargs):
body = kwargs.get('body')
bucket_id = body.get('bucket_id')
model_id = body.get('model_id')
features, target = get_data_by_bucket_id(bucket_id)
features = features.T
model_params_name = [
"weights_init", 'num_iters', 'batch_size', 'neighborhood',
'learning_rate', 'learning_decay_rate', 'sigma', 'sigma_decay_rate',
'num_clusters'
]
model_params = {
'model__{}'.format(p): v
for p, v in body.items() if p in model_params_name
}
print(model_params)
model_params['model__verbose'] = True
X_train = np.array(features)
y_train = np.array(target).astype(int)
try:
result_ml = helper.model_train(model_id, X_train, y_train,
**model_params)
except Exception as err:
config.logger.error(str(err))
# err_message = ml_models.result.ErrorResult()
return {"message": str(err)}, 400
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
def train(*args, **kwargs):
body = kwargs.get('body')
model_id = body.get('model_id')
features = body.get('features')
target = body.get('target')
model_params_name = [
'weights_init', 'num_iters', 'batch_size', 'neighborhood',
'learning_rate', 'learning_decay_rate', 'sigma', 'sigma_decay_rate',
'num_clusters'
]
model_params = {
'model__{}'.format(p): v
for p, v in body.items() if p in model_params_name
}
print(model_params)
model_params['model__verbose'] = True
X_train = np.array(features).T
y_train = np.array(target)
try:
result_ml = helper.model_train(model_id, X_train, y_train,
**model_params)
except Exception as err:
config.logger.error(str(err))
err_message = ml_models.result.ErrorResult()
return err_message()
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
def train_by_bucket_data(model_id, bucket_id):
features, target = get_data_by_bucket_id(bucket_id)
features = features.T
try:
result_ml = helper.model_train(model_id, features, target)
except Exception as err:
config.logger.error(str(err))
config.logger.error(traceback.print_exc())
# err_message = ml_models.result.ErrorResult()
return {"message": str(err)}, 400
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
def train(model_id, features, target):
features = np.array(features).T
target = np.array(target)
try:
result_ml = helper.model_train(model_id, features, target)
except Exception as err:
config.logger.error(str(err))
config.logger.error(traceback.print_exc())
err_message = ml_models.result.ErrorResult()
return err_message()
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
def train_by_bucket_data(model_id, bucket_id):
features, target = get_data_by_bucket_id(bucket_id)
print("-------------Feature-----------")
for i in features:
print(i)
print("================================")
print("--------------Target------------")
for i in target:
print(i)
print("================================")
features = features.T
try:
result_ml = helper.model_train(model_id, features, target)
except Exception as err:
config.logger.error(str(err))
config.logger.error(traceback.print_exc())
err_message = ml_models.result.ErrorResult()
return err_message()
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
def train_by_bucket_data(*args, **kwargs):
# features, target = get_data_by_bucket_id(bucket_id)
# features = features.T
# try:
# result_ml = helper.model_train(model_id, features, target)
# except Exception as err:
# config.logger.error(str(err))
# config.logger.error(traceback.print_exc())
# err_message = ml_models.result.ErrorResult()
# return err_message()
# else:
# success_message = ml_models.get_result(model_id)(**result_ml)
# return success_message()
body = kwargs.get('body')
bucket_id = body.get('bucket_id')
model_id = body.get('model_id')
features, target = get_data_by_bucket_id(bucket_id)
features = features.T
model_params_name = ["features_selection","subset_size", 'unsup_num_iters', 'unsup_batch_size', 'sup_num_iters', 'sup_batch_size',
'neighborhood', 'learning_rate', 'learning_decay_rate', 'sigma', 'sigma_decay_rate']
model_params = {'model__{}'.format(p): v for p, v in body.items() if p in model_params_name}
print(model_params)
model_params['model__verbose'] = True
X_train = np.array(features)
y_train = np.array(target).astype(int)
print(X_train.shape, y_train.shape)
try:
result_ml = helper.model_train(model_id, X_train, y_train, **model_params)
except Exception as err:
config.logger.error(str(err))
# err_message = ml_models.result.ErrorResult()
return {"message": str(err)}, 400
else:
success_message = ml_models.get_result(model_id)(**result_ml)
return success_message()
