dataclr.results

The dataclr.results module provides classes and structures to represent and manage the outputs of feature selection and model evaluation processes.

class dataclr.results.MethodResult(node)

This class provides a representation of the final result and the sequence of methods applied during the process. I

result

The final result of the feature selection or model evaluation.

Type:

Result

methods_list

A list of methods applied in the order they were executed.

Type:

list[Method]

Parameters:

node – A GraphNode object containing the result and its associated methods.

class dataclr.results.Result(params: dict[str, object], performance: ResultPerformance, feature_list: list[str])

Represents the result of a feature selection or model evaluation process.

params

The parameters used by the method to achieve this result.

Type:

dict[str, object]

performance

The performance metrics of the result.

Type:

ResultPerformance

feature_list

A list of selected features.

Type:

list[str]

class dataclr.results.ResultPerformance(r2: float = None, rmse: float = None, accuracy: float = None, precision: float = None, recall: float = None, f1: float = None)

Represents the performance metrics of a model or result.

This class serves as a base class for specific performance metrics, such as those for regression or classification tasks.

Subclasses:

• RegressionPerformance

• ClassificationPerformance

r2

Coefficient of determination (R²) score.

Type:

float

rmse

Root Mean Squared Error.

Type:

float

accuracy

Accuracy score.

Type:

float

precision

Precision score.

Type:

float

recall

Recall score.

Type:

float

f1

F1 score.

Type:

float