summaryrefslogtreecommitdiff
path: root/fuzzylite/fl/defuzzifier/WeightedAverageCustom.h
diff options
context:
space:
mode:
Diffstat (limited to 'fuzzylite/fl/defuzzifier/WeightedAverageCustom.h')
-rw-r--r--fuzzylite/fl/defuzzifier/WeightedAverageCustom.h77
1 files changed, 77 insertions, 0 deletions
diff --git a/fuzzylite/fl/defuzzifier/WeightedAverageCustom.h b/fuzzylite/fl/defuzzifier/WeightedAverageCustom.h
new file mode 100644
index 0000000..5597f97
--- /dev/null
+++ b/fuzzylite/fl/defuzzifier/WeightedAverageCustom.h
@@ -0,0 +1,77 @@
+/*
+ fuzzylite (R), a fuzzy logic control library in C++.
+ Copyright (C) 2010-2017 FuzzyLite Limited. All rights reserved.
+ Author: Juan Rada-Vilela, Ph.D. <jcrada@fuzzylite.com>
+
+ This file is part of fuzzylite.
+
+ fuzzylite is free software: you can redistribute it and/or modify it under
+ the terms of the FuzzyLite License included with the software.
+
+ You should have received a copy of the FuzzyLite License along with
+ fuzzylite. If not, see <http://www.fuzzylite.com/license/>.
+
+ fuzzylite is a registered trademark of FuzzyLite Limited.
+ */
+
+#ifndef FL_WEIGHTEDAVERAGECUSTOM_H
+#define FL_WEIGHTEDAVERAGECUSTOM_H
+
+#include "fl/defuzzifier/WeightedDefuzzifier.h"
+
+namespace fl {
+ class Activated;
+
+ /**
+ The (experimental) WeightedAverageCustom class is a WeightedDefuzzifier that computes the
+ weighted average of a fuzzy set represented in an Aggregated Term utilizing
+ the fuzzy operators for implication and aggregation to compute the weighted
+ average. This is an experimental approach to take advantage of customization
+ thanks to the object-oriented design.
+
+ @author Juan Rada-Vilela, Ph.D.
+ @see WeightedAverage
+ @see WeightedSum
+ @see WeightedSumCustom
+ @see WeightedDefuzzifier
+ @see Defuzzifier
+ @since 6.0
+ */
+ class FL_API WeightedAverageCustom : public WeightedDefuzzifier {
+ public:
+ explicit WeightedAverageCustom(Type type = Automatic);
+ explicit WeightedAverageCustom(const std::string& type);
+ virtual ~WeightedAverageCustom() FL_IOVERRIDE;
+ FL_DEFAULT_COPY_AND_MOVE(WeightedAverageCustom)
+
+ virtual std::string className() const FL_IOVERRIDE;
+
+ virtual Complexity complexity(const Term* term) const FL_IOVERRIDE;
+
+ /**
+ Computes the weighted average of the given fuzzy set represented as
+ an AggregatedTerm as @f$y = \dfrac{\sum_i w_iz_i}{\sum_i w_i} @f$,
+ where @f$w_i@f$ is the activation degree of term @f$i@f$, and
+ @f$z_i = \mu_i(w_i) @f$.
+
+ If the implication and aggregation operators are set to fl::null (or
+ set to AlgebraicProduct and UnboundedSum, respectively), then the
+ operation of WeightedAverageCustom is the same as the WeightedAverage.
+ Otherwise, the implication and aggregation operators are utilized to
+ compute the multiplications and sums in @f$y@f$, respectively.
+
+ @param term is the fuzzy set represented as an Aggregated Term
+ @param minimum is the minimum value of the range (only used for Tsukamoto)
+ @param maximum is the maximum value of the range (only used for Tsukamoto)
+ @return the weighted average of the given fuzzy set
+ */
+ virtual scalar defuzzify(const Term* term,
+ scalar minimum, scalar maximum) const FL_IOVERRIDE;
+ virtual WeightedAverageCustom* clone() const FL_IOVERRIDE;
+
+ static Defuzzifier* constructor();
+ };
+}
+
+#endif /* FL_WEIGHTEDAVERAGECUSTOM_H */
+