iris_classification_lambda / inference_direct.py
Clement Vachet
Improve code based on pylint and black suggestions
9ecca49
raw
history blame
667 Bytes
"""
Direct inference with hard-coded data
"""
import json
from classification.classifier import Classifier
if __name__ == "__main__":
cls = Classifier()
# Training
cls.train_and_save()
# Testing
data = {"features": [[6.5, 3.0, 5.8, 2.2], [6.1, 2.8, 4.7, 1.2]]}
features = data["features"]
results = cls.load_and_test(features)
print("results:", results)
# Response similar to REST API call
response = {
"statusCode": 200,
"body": json.dumps(
{"predictions": results["predictions"], "probabilities": results["probabilities"]}
),
}
print("Example REST API response: ", response)