[yul-phaser] Add Population class

tools/CMakeLists.txt

@@ -15,6 +15,8 @@ install(TARGETS solidity-upgrade DESTINATION "${CMAKE_INSTALL_BINDIR}")
 
 add_executable(yul-phaser
 	yulPhaser/main.cpp
+	yulPhaser/Population.h
+	yulPhaser/Population.cpp
 	yulPhaser/Chromosome.h
 	yulPhaser/Chromosome.cpp
 	yulPhaser/Program.h

tools/yulPhaser/Population.cpp

@@ -0,0 +1,137 @@
+/*
+	This file is part of solidity.
+
+	solidity is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	solidity is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with solidity.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#include <tools/yulPhaser/Population.h>
+
+#include <tools/yulPhaser/Program.h>
+
+#include <algorithm>
+#include <cassert>
+#include <iostream>
+#include <numeric>
+
+using namespace std;
+using namespace solidity;
+using namespace solidity::phaser;
+
+namespace solidity::phaser
+{
+
+ostream& operator<<(ostream& _stream, Individual const& _individual);
+ostream& operator<<(ostream& _stream, Population const& _population);
+
+}
+
+ostream& phaser::operator<<(ostream& _stream, Individual const& _individual)
+{
+	_stream << "Fitness: ";
+	if (_individual.fitness.has_value())
+		_stream << _individual.fitness.value();
+	else
+		_stream << "<NONE>";
+	_stream << ", optimisations: " << _individual.chromosome;
+
+	return _stream;
+}
+
+Population::Population(string const& _sourcePath, vector<Chromosome> const& _chromosomes):
+	m_sourcePath{_sourcePath}
+{
+	for (auto const& chromosome: _chromosomes)
+		m_individuals.push_back({chromosome});
+}
+
+Population Population::makeRandom(string const& _sourcePath, size_t _size)
+{
+	vector<Individual> individuals;
+	for (size_t i = 0; i < _size; ++i)
+		individuals.push_back({Chromosome::makeRandom(randomChromosomeLength())});
+
+	return Population(_sourcePath, individuals);
+}
+
+size_t Population::measureFitness(Chromosome const& _chromosome, string const& _sourcePath)
+{
+	auto program = Program::load(_sourcePath);
+	program.optimise(_chromosome.optimisationSteps());
+	return program.codeSize();
+}
+
+void Population::run(optional<size_t> _numRounds, ostream& _outputStream)
+{
+	doEvaluation();
+	for (size_t round = 0; !_numRounds.has_value() || round < _numRounds.value(); ++round)
+	{
+		doMutation();
+		doSelection();
+		doEvaluation();
+
+		_outputStream << "---------- ROUND " << round << " ----------" << endl;
+		_outputStream << *this;
+	}
+}
+
+ostream& phaser::operator<<(ostream& _stream, Population const& _population)
+{
+	_stream << "Source: " << _population.m_sourcePath << endl;
+
+	auto individual = _population.m_individuals.begin();
+	for (; individual != _population.m_individuals.end(); ++individual)
+		_stream << *individual << endl;
+
+	return _stream;
+}
+
+void Population::doMutation()
+{
+	// TODO: Implement mutation and crossover
+}
+
+void Population::doEvaluation()
+{
+	for (auto& individual: m_individuals)
+		if (!individual.fitness.has_value())
+			individual.fitness = measureFitness(individual.chromosome, m_sourcePath);
+}
+
+void Population::doSelection()
+{
+	assert(all_of(m_individuals.begin(), m_individuals.end(), [](auto& i){ return i.fitness.has_value(); }));
+
+	sort(
+		m_individuals.begin(),
+		m_individuals.end(),
+		[](auto const& a, auto const& b){ return a.fitness.value() < b.fitness.value(); }
+	);
+
+	randomizeWorstChromosomes(m_individuals, m_individuals.size() / 2);
+}
+
+void Population::randomizeWorstChromosomes(
+	vector<Individual>& _individuals,
+	size_t _count
+)
+{
+	assert(_individuals.size() >= _count);
+	// ASSUMPTION: _individuals is sorted in ascending order
+
+	auto individual = _individuals.begin() + (_individuals.size() - _count);
+	for (; individual != _individuals.end(); ++individual)
+	{
+		*individual = {Chromosome::makeRandom(randomChromosomeLength())};
+	}
+}

tools/yulPhaser/Population.h

@@ -0,0 +1,87 @@
+/*
+	This file is part of solidity.
+
+	solidity is free software: you can redistribute it and/or modify
+	it under the terms of the GNU General Public License as published by
+	the Free Software Foundation, either version 3 of the License, or
+	(at your option) any later version.
+
+	solidity is distributed in the hope that it will be useful,
+	but WITHOUT ANY WARRANTY; without even the implied warranty of
+	MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+	GNU General Public License for more details.
+
+	You should have received a copy of the GNU General Public License
+	along with solidity.  If not, see <http://www.gnu.org/licenses/>.
+*/
+
+#pragma once
+
+#include <tools/yulPhaser/Chromosome.h>
+#include <tools/yulPhaser/Random.h>
+
+#include <optional>
+#include <ostream>
+#include <vector>
+
+namespace solidity::phaser
+{
+
+/**
+ * Information describing the state of an individual member of the population during the course
+ * of the genetic algorithm.
+ */
+struct Individual
+{
+	Chromosome chromosome;
+	std::optional<size_t> fitness = std::nullopt;
+
+	friend std::ostream& operator<<(std::ostream& _stream, Individual const& _individual);
+};
+
+/**
+ * Represents a changing set of individuals undergoing a genetic algorithm.
+ * Each round of the algorithm involves mutating existing individuals, evaluating their fitness
+ * and selecting the best ones for the next round.
+ *
+ * An individual is a sequence of optimiser steps represented by a @a Chromosome instance. The whole
+ * population is associated with a fixed Yul program. By loading the source code into a @a Program
+ * instance the class can compute fitness of the individual.
+ */
+class Population
+{
+public:
+	static constexpr size_t MaxChromosomeLength = 30;
+
+	explicit Population(std::string const& _sourcePath, std::vector<Chromosome> const& _chromosomes = {});
+	static Population makeRandom(std::string const& _sourcePath, size_t _size);
+
+	void run(std::optional<size_t> _numRounds, std::ostream& _outputStream);
+
+	std::vector<Individual> const& individuals() const { return m_individuals; }
+
+	static size_t randomChromosomeLength() { return binomialRandomInt(MaxChromosomeLength, 0.5); }
+	static size_t measureFitness(Chromosome const& _chromosome, std::string const& _sourcePath);
+
+	friend std::ostream& operator<<(std::ostream& _stream, Population const& _population);
+
+private:
+	explicit Population(std::string const& _sourcePath, std::vector<Individual> _individuals = {}):
+		m_sourcePath{_sourcePath},
+		m_individuals{std::move(_individuals)} {}
+
+	void doMutation();
+	void doEvaluation();
+	void doSelection();
+
+	static void randomizeWorstChromosomes(
+		std::vector<Individual>& _individuals,
+		size_t _count
+	);
+
+	std::string m_sourcePath;
+
+	std::vector<Individual> m_individuals;
+};
+
+}

tools/yulPhaser/main.cpp

@@ -15,9 +15,8 @@
 	along with solidity.  If not, see <http://www.gnu.org/licenses/>.
 */
 
-#include <tools/yulPhaser/Chromosome.h>
 #include <tools/yulPhaser/Exceptions.h>
-#include <tools/yulPhaser/Program.h>
+#include <tools/yulPhaser/Population.h>
 
 #include <boost/program_options.hpp>
 
@@ -40,8 +39,8 @@ struct CommandLineParsingResult
 
 void runAlgorithm(string const& _sourcePath)
 {
-	Program::load(_sourcePath).optimize(Chromosome::makeRandom(15).asSequence());
-	cout << "Program load and optimization successful." << endl;
+	auto population = Population::makeRandom(_sourcePath, 10);
+	population.run(nullopt, cout);
 }
 
 CommandLineParsingResult parseCommandLine(int argc, char** argv)