In an attempt to better understand this framework, I've added some documentation to the anomalyDetector.py file, as I noticed that in the model architecture page it notes that every function should have a docstring. I did my best to follow the style outlined on the architecture page and from observing other files.
A couple of small things I noticed while digging into and trying to understand what's going on in this notebook:
-In the workProphet function, this appears to run three separate models, but the function name seems to imply that it's only for the first model, The Prophet monovariate anomaly detector.
-Both the workLof and workIso functions contain input parameter modelDir, but don't use it
dpinney
commented
4 years ago
In an attempt to better understand this framework, I've added some documentation to the anomalyDetector.py file, as I noticed that in the model architecture page it notes that every function should have a docstring. I did my best to follow the style outlined on the architecture page and from observing other files.
A couple of small things I noticed while digging into and trying to understand what's going on in this notebook: -In the workProphet function, this appears to run three separate models, but the function name seems to imply that it's only for the first model, The Prophet monovariate anomaly detector. -Both the workLof and workIso functions contain input parameter modelDir, but don't use it