make_monitoring_ui_artifacts module

make_monitoring_ui_artifacts.load_data_from_db() → tuple[DataFrame, DataFrame]

Load dataset with meaningful features, and same dataset but turned into dummies (dataset consists of only categorical features)

make_monitoring_ui_artifacts.load_model_from_mlflow(model_mlflow_runid: str = None) → object

Load a model to be used If a run id is not provided, the env run id will be used

make_monitoring_ui_artifacts.make_evidently_html_dashboards(meaningful_reference_data: DataFrame, meaningful_current_data: DataFrame) → None

Create HTML evidently dashboards

make_monitoring_ui_artifacts.make_monitoring_ui_artifacts()

Update monitoring UI artifacts used by streamlit Default (from env) model run id is used, the user can input a new mlflow run id to use a new model

make_monitoring_ui_artifacts.make_shap_graphs(model: object, X: DataFrame) → None

Make SHAP graphs based on loaded model and data used for model training

make_monitoring_ui_artifacts.prep_data_for_shap_graphs(model_data_w_dummy: DataFrame) → DataFrame

Prepare data for the next task in the flow

make_monitoring_ui_artifacts.prepare_data_for_evidently(model_data_w_dummy: DataFrame, meaningful_features_data: DataFrame) → tuple[DataFrame, DataFrame]

I only have 1 set of data, so I split it to create reference/current data