Commit bd353ed9 authored by dasharatha.vamshi's avatar dasharatha.vamshi

pkl

parent f06b4bd0
......@@ -41,10 +41,11 @@ class PreProcessComponent:
sc_X = self.read_pickle_file(standard_scalar_pkl_path)
X = sc_X.transform(forcast_data)
return X
return X, self.date
if __name__ == '__main__':
logger.info(config)
# Checking shared Volume
if PreProcessConstants.SHARED_VOLUME in config.keys():
shared_volume = config[PreProcessConstants.SHARED_VOLUME]
......@@ -60,6 +61,10 @@ if __name__ == '__main__':
# Checking pickle path for standard scalar
if PreProcessConstants.STANDARD_SCALAR_PATH in config.keys():
standard_scalar_path = config[PreProcessConstants.STANDARD_SCALAR_PATH]
if standard_scalar_path is None:
standard_scalar_path = "StandardScaler.pkl"
else:
pass
else:
raise Exception(ComponentExceptions.INVALID_JSON_PATH)
......@@ -68,14 +73,16 @@ if __name__ == '__main__':
fillna_method = config[PreProcessConstants.FILLNA_METHOD]
else:
raise Exception(ComponentExceptions.INVALID_Fillna_Method)
obj = PreProcessComponent()
data = obj.preprocess(json_path, standard_scalar_path, fillna_method)
logger.info("Got the data writing it to pickle file (preprocessed_X.pkl)")
data, date_pkl = obj.preprocess(json_path, standard_scalar_path, fillna_method)
logger.info("Got the data writing it to pickle files (preprocessed_X.pkl and date_X.pkl)")
try:
output = open(os.path.join(shared_volume, 'preprocessed_X.pkl'), 'wb')
pickle.dump(data, output)
output.close()
output1 = open(os.path.join(shared_volume, 'date_X.pkl'), 'wb')
pickle.dump(date_pkl, output1)
output1.close()
logger.info("Component executed Successfully")
except Exception as e:
raise Exception(e)
......@@ -30,6 +30,7 @@ LOGSTASH_PORT = str(_config.get('SERVICE_CONFIG', {}).get('LOGSTASH_PORT'))
# os.environ["json_path"] = r"E:\welspun-defects\preprocess_data-ilendev\response.json"
# os.environ["fillna_method"] = "ffill"
# os.environ["standard_scalar_path"] = r"E:\welspun-defects\preprocess_data-ilendev\StandardScaler.pkl"
print(os.getcwd())
PKL_path = os.path.join(os.getcwd(), 'StandardScaler.pkl')
config = {
"shared_volume": os.environ.get("shared_volume"),
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment