require 'svmkit/base/evaluator' module SVMKit # This module consists of the classes for model evaluation. module EvaluationMeasure # Accuracy is a class that calculates the accuracy of classifier from the predicted labels. # # @example # evaluator = SVMKit::EvaluationMeasure::Accuracy.new # puts evaluator.score(ground_truth, predicted) class Accuracy include Base::Evaluator # Claculate mean accuracy. # # @param y_true [Numo::Int32] (shape: [n_samples]) Ground truth labels. # @param y_pred [Numo::Int32] (shape: [n_samples]) Predicted labels. # @return [Float] Mean accuracy def score(y_true, y_pred) (y_true.to_a.map.with_index { |label, n| label == y_pred[n] ? 1 : 0 }).inject(:+) / y_true.size.to_f end end end end