Metrics.hpp File Reference

Mode-specific evaluation metric blocks. More...

#include "stats/ConfusionMatrix.hpp"
#include "utils/Types.hpp"
#include <optional>
#include <variant>

Go to the source code of this file.

Classes
struct	ppforest2::stats::ClassificationMetrics
	Classification evaluation metrics. More...
struct	ppforest2::stats::RegressionMetrics
	Regression evaluation metrics. More...

Namespaces
namespace	ppforest2
	Binarization strategies for multiclass-to-binary reduction.
namespace	ppforest2::stats
	Statistical infrastructure for training and evaluation.

Typedefs
using	ppforest2::stats::Metrics = std::variant<ClassificationMetrics, RegressionMetrics>
	Mode-polymorphic metrics block.

Functions
float	ppforest2::stats::accuracy (types::OutcomeVector const &predictions, types::GroupIdVector const &actual)
	Fraction of predictions matching the ground-truth class label.
double	ppforest2::stats::error_rate (types::OutcomeVector const &predictions, types::GroupIdVector const &actual)
	Misclassification rate — 1 - accuracy.
double	ppforest2::stats::error_rate (types::OutcomeVector const &predictions, types::OutcomeVector const &actual)
	Convenience overload: float-typed labels (cast to GroupId locally).
double	ppforest2::stats::mae (types::OutcomeVector const &predictions, types::OutcomeVector const &actual)
	Mean absolute error.
Metrics	ppforest2::stats::metrics_from_outcomes (types::OutcomeVector const &y_pred, types::OutcomeVector const &y, types::Mode mode)
	Build a mode-appropriate Metrics variant from in-memory tensors.
double	ppforest2::stats::mse (types::OutcomeVector const &predictions, types::OutcomeVector const &actual)
	Mean squared error.
double	ppforest2::stats::r_squared (types::OutcomeVector const &predictions, types::OutcomeVector const &actual)
	Coefficient of determination (R²). Returns 0 when total variance is 0.

Detailed Description

Mode-specific evaluation metric blocks.

Pairs ClassificationMetrics (confusion matrix + error rate) with RegressionMetrics (MSE / MAE / R²). Both are consumed together by the OOB / variant-based code paths in models/Evaluation.hpp, serialization, and presentation, so they live side by side.