iguanas.rule_scoring
.RuleScorer¶
- class iguanas.rule_scoring.RuleScorer(scoring_class: Union[iguanas.rule_scoring.rule_scoring_methods.PerformanceScorer, iguanas.rule_scoring.rule_scoring_methods.LogRegScorer, iguanas.rule_scoring.rule_scoring_methods.RandomForestScorer], scaling_class=None)[source]¶
Generates rule scores using the rule binary columns and the target column.
- Parameters
- scoring_classUnion[PerformanceScorer, LogRegScorer, RandomForestScorer]
The instantiated scoring class - this defines the method for generating the scores. Scoring classes are available in the rule_scoring_methods module.
- scaling_classUnion[MinMaxScaler, ConstantScaler], optional
The instantiated scaling class - this defines the method for scaling the raw scores from the scoring class. Scaling classes are available in the rule_score_scalers module. Defaults to None.
- Attributes
- rule_scoresDict[str, int]
Contains the generated score (values) for each rule (keys).
- fit(X_rules: iguanas.utils.typing.pandas.core.frame.DataFrame, y: iguanas.utils.typing.pandas.core.series.Series, sample_weight=None) None [source]¶
Generates rule scores using the rule binary columns and the binary target column.
- Parameters
- X_rulesPandasDataFrameType
The rule binary columns.
- yPandasSeriesType
The binary target column.
- sample_weightPandasSeriesType, optional
Row-wise weights to apply in the scoring_class. Defaults to None.
- transform(X_rules: iguanas.utils.typing.pandas.core.frame.DataFrame) iguanas.utils.typing.pandas.core.frame.DataFrame [source]¶
Transforms the rule binary columns to show the generated scores applied to the dataset (i.e. replaces the 1 in X_rules with the generated score).
- Parameters
- X_rulesPandasDataFrameType
The rule binary columns.
- Returns
- PandasDataFrameType
The generated scores applied to the dataset.
- fit_transform(X_rules: iguanas.utils.typing.pandas.core.frame.DataFrame, y: iguanas.utils.typing.pandas.core.series.Series, sample_weight=None) iguanas.utils.typing.pandas.core.frame.DataFrame [source]¶
Generates rule scores using the rule binary columns and the binary target column, then transforms the rule binary columns to show the generated scores applied to the dataset (i.e. replaces the 1 in X_rules with the generated score).
- Parameters
- X_rulesPandasDataFrameType
The rule binary columns.
- yPandasSeriesType
The binary target column.
- sample_weightPandasSeriesType, optional
Row-wise weights to apply in the scoring_class. Defaults to None.
- Returns
- PandasDataFrameType
The generated scores applied to the dataset.