iguanas.metrics.unsupervised
.AlertsPerDay¶
- class iguanas.metrics.unsupervised.AlertsPerDay(n_alerts_expected_per_day: int, no_of_days_in_file: int)[source]¶
Calculates the negative squared difference between the number of alerts per day in the binary predictor(s) vs the expected.
- Parameters
- n_alerts_expected_per_dayint
Expected number of alerts for the given rule.
- no_of_days_in_fileint
Number of days of data provided in the file.
- fit(y_preds: Union[iguanas.utils.typing.numpy.ndarray, iguanas.utils.typing.pandas.core.series.Series, iguanas.utils.typing.databricks.koalas.series.Series, iguanas.utils.typing.pandas.core.frame.DataFrame, iguanas.utils.typing.databricks.koalas.frame.DataFrame]) Union[float, iguanas.utils.typing.numpy.ndarray] [source]¶
Calculates the negative squared difference between the number of alerts per day in the binary predictor(s) vs the expected.
- Parameters
- y_predsUnion[NumpyArrayType, PandasSeriesType, KoalasSeriesType, PandasDataFrameType, KoalasDataFrameType]
The binary predictor column(s).
- Returns
- Union[float, NumpyArrayType]
The negative squared difference(s).