Model Persistence

Model persistence is helpful for saving and loading BTYD models for reuse as well as logging to a registry server such as mlflow. After a model has been fit, its inference data can be saved in either JSON or CSV format. The format will be inferred automatically from the filename.

Fit model

from btyd import BetaGeoModel
from btyd.datasets import load_cdnow_summary

data = load_cdnow_summary(index_col=[0])
bgm = BetaGeoModel()
bgm.fit(data)
bgm
"""<btyd.BetaGeoModel: Parameters {'alpha': 4.5, 'r': 0.2, 'a': 0.8, 'b': 2.4} estimated with 2357 customers.>"""

Saving and Loading Models

# Save inference data of a fitted model as a JSON:
bgm.save('path/to/file.json')

# Load model inference data from a JSON into a new or existing model:
bgm = bgm.load('path/to/file.json')
bgm_new = BetaGeoModel()
bgm_new._idata = bgm.load('path/to/file.json')

bgm_new
"""<btyd.BetaGeoModel: Parameters {'alpha': 4.5, 'r': 0.2, 'a': 0.8, 'b': 2.4} estimated with 2357 customers.>"""