*** empty log message ***

This commit is contained in:
mac 2000-02-16 15:11:18 +00:00
commit 4ebd212d14
15 changed files with 1872 additions and 452 deletions

54
eo/src/eoBackInserter.h Normal file
View file

@ -0,0 +1,54 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoInserter.h
Abstract population insertion operator, which is used by the eoGeneralOps
to insert the results in the (intermediate) population. This file also
contains the definitions of a derived classes that implements a back inserter,
probably the only efficient inserter for populations of type vector.
(c) Maarten Keijzer (mkeijzer@mad.scientist.com) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoBackInserter_h
#define eoBackInserter_h
#include "eoInserter.h"
/**
* eoBackInserter: Interface class that enables an operator to insert
new individuals at the back of the new population.
*/
template <class EOT>
class eoBackInserter : public eoPopInserter<EOT>
{
public :
eoBackInserter(void) : eoPopInserter<EOT>() {}
void insert(const EOT& _eot)
{
pop().push_back(_eot);
}
string className(void) const { return "eoBackInserter"; }
};
#endif

View file

@ -0,0 +1,57 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoDetTournamentIndiSelector.h
(c) Maarten Keijzer (mkeijzer@mad.scientist.com) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoDetTournamentIndiSelector_h
#define eoDetTournamentIndiSelector_h
#include "eoIndiSelector.h"
#include "selectors.h"
/**
* eoDetTournamentIndiSelector: selects children through a deterministic_tournament
*/
template <class EOT>
class eoDetTournamentIndiSelector : public eoPopIndiSelector<EOT>
{
public :
eoDetTournamentIndiSelector(int _tournamentSize)
: eoPopIndiSelector<EOT>(),
tournamentSize(_tournamentSize)
{}
virtual ~eoDetTournamentIndiSelector(void) {}
const EOT& do_select(void)
{
return *deterministic_tournament(begin(), end(), tournamentSize);
}
private :
int tournamentSize;
};
#endif

134
eo/src/eoIndiSelector.h Normal file
View file

@ -0,0 +1,134 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoIndiSelector.h
Abstract selection operator, which is used by the eoGeneralOps
to obtain individuals from a source population. It also gives a
direct descended eoPopIndiSelector that can be used to
initialize objects with an eoPop<EOT>. For most uses use eoPopIndividualSelector
rather than eoIndividualSelector to derive from.
(c) Maarten Keijzer (mkeijzer@mad.scientist.com) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoIndiSelector_h
#define eoIndiSelector_h
/**
* eoIndividualSelector: This class defines the interface
*/
template <class EOT>
class eoIndiSelector
{
public :
eoIndiSelector() {}
virtual ~eoIndiSelector(void) {}
virtual size_t size(void) const = 0;
virtual const EOT& operator[](size_t) const = 0;
virtual const EOT& select(void) = 0;
virtual vector<const EOT*> select(size_t _how_many)
{ // default implementation just calls select a couple of times
// this can be overridden in favour of a more efficient implementation
vector<const EOT*> result(_how_many);
for (int i = 0; i < _how_many; ++i)
{
result[i] = &select();
}
return result;
}
};
/**
* eoPopIndiSelector: Intermediate class for dispensing populations
various useful things can be done with this class:
you can specify how many of the population can ever be dispensed to the
operators, but you can also specify a preference to the first guy being
dispensed. This is useful if you want to perform the operator on a specific
individual.
*/
template <class EOT>
class eoPopIndiSelector : public eoIndiSelector<EOT>
{
public :
eoPopIndiSelector(void) : pop(0), firstChoice(-1), last(0), eoIndiSelector<EOT>() {}
virtual ~eoPopIndiSelector(void) {}
struct eoUnitializedException{};
/** Initialization function
*/
eoPopIndiSelector& operator()(const eoPop<EOT>& _pop, int _end = -1, int _myGuy = -1)
{
pop = &_pop;
last = _end;
if (last < 0 || last > pop->size())
{
last = pop->size();
}
firstChoice = _myGuy;
return *this;
}
size_t size(void) const { valid(); return last; }
const EOT& operator[](size_t _i) const { valid(); return pop->operator[](_i); }
eoPop<EOT>::const_iterator begin(void) const { valid(); return pop->begin(); }
eoPop<EOT>::const_iterator end(void) const { valid(); return pop->end(); }
/// select does the work. Note that it is not virtual. It calls do_select that needs to be implemented by the derived classes
const EOT& select(void)
{
valid();
if (firstChoice < 0 || firstChoice >= size())
{
return do_select(); // let the child figure out what to do
}
const EOT& result = pop->operator[](firstChoice);
firstChoice = -1;
return result;
}
virtual const EOT& do_select(void) = 0;
private :
void valid(void) const
{
if (pop == 0)
throw eoUnitializedException();
}
const eoPop<EOT>* pop; // need a pointer as this the pop argument can be re-instated
int last;
int firstChoice;
};
#endif

93
eo/src/eoInserter.h Normal file
View file

@ -0,0 +1,93 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoInserter.h
Abstract population insertion operator, which is used by the eoGeneralOps
to insert the results in the (intermediate) population. It also contains
a direct descended eoPopInserter that defines a convenient inbetween class
for working with eoPop<EOT>. The user will most likely derive from eoPopInserter
rather than eoInserter.
(c) Maarten Keijzer (mak@dhi.dk) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoInserter_h
#define eoInserter_h
#include "eoObject.h"
/**
* eoInserter: Interface class that enables an operator to insert
new individuals into the (intermediate) population.
*/
template <class EOT>
class eoInserter : public eoObject
{
public :
virtual ~eoInserter() {}
struct eoInserterException{};
virtual void insert(const EOT&) = 0; // can throw an eoInserterException
};
/**
* eoPopInserter: In-between class that defines an initialization
* of the eoIndividualInserter.
*/
template <class EOT>
class eoPopInserter : public eoInserter<EOT>
{
public :
eoPopInserter(void) : thePop(0), eoInserter<EOT>() {}
/// Binds the population to this class. This is an initialization routine used by breeders
eoInserter<EOT>& operator()(eoPop<EOT>& _pop)
{
thePop = &_pop;
return *this;
}
protected :
eoPop<EOT>& pop(void) const { valid(); return *thePop; }
private :
void valid(void) const
{
if (thePop == 0)
throw eoInserterException();
}
// Need a pointer as the inserter should be able to bind to different populations.
// This is caused by the 'one template parameter only' convention in EO.
eoPop<EOT>* thePop;
// If eoGOpBreeder could be templatized over the inserter and the selector,
// the pop could be a ref as this class could be created every time it is applied
// and subsequently would get the population through the constructor
};
#endif

View file

@ -0,0 +1,54 @@
/** -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoProportionalGOpSelector.h
Proportional Generalized Operator Selector.
(c) Maarten Keijzer (mkeijzer@mad.scientist.com), GeNeura Team 1998, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoProportionalGOpSelector_h
#define eoProportionalGOpSelector_h
//-----------------------------------------------------------------------------
#include "eoGOpSelector.h"
/** eoProportionalGOpSel: do proportional selection, returns one of the
operators
*/
template <class EOT>
class eoProportionalGOpSel : public eoGOpSelector<EOT>
{
public :
eoProportionalGOpSel() : eoGOpSelector<EOT>() {}
/** Returns the operator proportionally selected */
virtual eoGeneralOp<EOT>& selectOp()
{
unsigned what = rng.roulette_wheel(getRates());
return *operator[](what);
}
///
virtual string className() const { return "eoGOpSelector"; };
};
#endif

View file

@ -1,452 +1,452 @@
/*
* Random number generator adapted from (see comments below)
*
* The random number generator is modified into a class
* by Maarten Keijzer (mak@dhi.dk). Also added the Box-Muller
* transformation to generate normal deviates.
*
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
/* ************ DOCUMENTATION IN ORIGINAL FILE *********************/
// This is the ``Mersenne Twister'' random number generator MT19937, which
// generates pseudorandom integers uniformly distributed in 0..(2^32 - 1)
// starting from any odd seed in 0..(2^32 - 1). This version is a recode
// by Shawn Cokus (Cokus@math.washington.edu) on March 8, 1998 of a version by
// Takuji Nishimura (who had suggestions from Topher Cooper and Marc Rieffel in
// July-August 1997).
//
// Effectiveness of the recoding (on Goedel2.math.washington.edu, a DEC Alpha
// running OSF/1) using GCC -O3 as a compiler: before recoding: 51.6 sec. to
// generate 300 million random numbers; after recoding: 24.0 sec. for the same
// (i.e., 46.5% of original time), so speed is now about 12.5 million random
// number generations per second on this machine.
//
// According to the URL <http://www.math.keio.ac.jp/~matumoto/emt.html>
// (and paraphrasing a bit in places), the Mersenne Twister is ``designed
// with consideration of the flaws of various existing generators,'' has
// a period of 2^19937 - 1, gives a sequence that is 623-dimensionally
// equidistributed, and ``has passed many stringent tests, including the
// die-hard test of G. Marsaglia and the load test of P. Hellekalek and
// S. Wegenkittl.'' It is efficient in memory usage (typically using 2506
// to 5012 bytes of static data, depending on data type sizes, and the code
// is quite short as well). It generates random numbers in batches of 624
// at a time, so the caching and pipelining of modern systems is exploited.
// It is also divide- and mod-free.
//
// This library is free software; you can redistribute it and/or modify it
// under the terms of the GNU Library General Public License as published by
// the Free Software Foundation (either version 2 of the License or, at your
// option, any later version). This library 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 Library General Public License for more details. You should have
// received a copy of the GNU Library General Public License along with this
// library; if not, write to the Free Software Foundation, Inc., 59 Temple
// Place, Suite 330, Boston, MA 02111-1307, USA.
//
// The code as Shawn received it included the following notice:
//
// Copyright (C) 1997 Makoto Matsumoto and Takuji Nishimura. When
// you use this, send an e-mail to <matumoto@math.keio.ac.jp> with
// an appropriate reference to your work.
//
// It would be nice to CC: <Cokus@math.washington.edu> when you write.
//
//
// uint32 must be an unsigned integer type capable of holding at least 32
// bits; exactly 32 should be fastest, but 64 is better on an Alpha with
// GCC at -O3 optimization so try your options and see what's best for you
//
/* ************ END DOCUMENTATION IN ORIGINAL FILE *********************/
#ifndef EO_RANDOM_NUMBER_GENERATOR
#define EO_RANDOM_NUMBER_GENERATOR
#include <ctime>
#include <eoPersistent.h>
#include <eoObject.h>
// TODO: check for various compilers if this is exactly 32 bits
// Unfortunately MSVC's preprocessor does not comprehends sizeof()
// so neat preprocessing tricks will not work
typedef unsigned long uint32; // Compiler and platform dependent!
//-----------------------------------------------------------------------------
// eoRng
//-----------------------------------------------------------------------------
/**
eoRng is a persitent class that uses the ``Mersenne Twister'' random number generator MT19937
for generating random numbers. The various member functions implement useful functions
for evolutionary algorithms. Included are: rand(), random(), flip() and normal().
Note for people porting EO to other platforms: please make sure that the typedef
uint32 in the file eoRng.h is exactly 32 bits long. It may be longer, but not
shorter. If it is longer, file compatibility between EO on different platforms
may be broken.
*/
class eoRng : public eoObject, public eoPersistent
{
public :
/**
ctor takes a random seed; if you want another seed, use reseed
@see reseed
*/
eoRng(uint32 s = (uint32) time(0) ) : state(0), next(0), left(-1), cached(false), N(624), M(397), K(0x9908B0DFU) {
state = new uint32[N+1];
initialize(s);
}
~eoRng(void)
{
delete [] state;
}
/**
Re-initializes the Random Number Generator.
*/
void reseed(uint32 s)
{
initialize(s);
}
/**
uniform(m = 1.0) returns a random double in the range [0, m)
*/
double uniform(double m = 1.0)
{ // random number between [0, m]
return m * double(rand()) / double(rand_max());
}
/**
random() returns a random integer in the range [0, m)
*/
uint32 random(uint32 m)
{
return uint32(uniform() * double(m));
}
/**
flip() tosses a biased coin such that flip(x/100.0) will
returns true x% of the time
*/
bool flip(float bias)
{
return uniform() < bias;
}
/**
normal() zero mean gaussian deviate with standard deviation of 1
*/
double normal(void); // gaussian mutation, stdev 1
/**
normal(stdev) zero mean gaussian deviate with user defined standard deviation
*/
double normal(double stdev)
{
return stdev * normal();
}
/**
normal(mean, stdev) user defined mean gaussian deviate with user defined standard deviation
*/
double normal(double mean, double stdev)
{
return mean + normal(stdev);
}
/**
rand() returns a random number in the range [0, rand_max)
*/
uint32 rand();
/**
rand_max() the maximum returned by rand()
*/
uint32 rand_max(void) const { return (uint32) 0xffffffff; }
/**
roulette_wheel(vec, total = 0) does a roulette wheel selection
on the input vector vec. If the total is not supplied, it is
calculated. It returns an integer denoting the selected argument.
*/
template <class T>
int roulette_wheel(const std::vector<T>& vec, T total = 0)
{
if (total == 0)
{ // count
for (unsigned i = 0; i < vec.size(); ++i)
total += vec[i];
}
float change = uniform() * total;
int i = 0;
while (change > 0)
{
change -= vec[i++];
}
return --i;
}
///
void printOn(ostream& _os) const
{
for (int i = 0; i < N; ++i)
{
_os << state[i] << ' ';
}
_os << int(next - state) << ' ';
_os << left << ' ' << cached << ' ' << cacheValue;
}
///
void readFrom(istream& _is)
{
for (int i = 0; i < N; ++i)
{
_is >> state[i];
}
int n;
_is >> n;
next = state + n;
_is >> left;
_is >> cached;
_is >> cacheValue;
}
private :
uint32 restart(void);
void initialize(uint32 seed);
uint32* state; // the array for the state
uint32* next;
int left;
bool cached;
float cacheValue;
const int N;
const int M;
const uint32 K; // a magic constant
/**
Private copy ctor and assignment operator to make sure that
nobody accidentally copies the random number generator.
If you want similar RNG's, make two RNG's and initialize
them with the same seed.
*/
eoRng (const eoRng&); // no implementation
eoRng& operator=(const eoRng&); // dito
};
/**
The one and only global eoRng object
*/
static eoRng rng;
/**
The class uniform_generator can be used in the STL generate function
to easily generate random floats and doubles between [0, _max). _max
defaults to 1.0
*/
template <class T = double> class uniform_generator
{
public :
uniform_generator(T _max = T(1.0), eoRng& _rng = rng) : maxim(_max), uniform(_rng) {}
virtual T operator()(void) { return (T) uniform.uniform(maxim); }
private :
T maxim;
eoRng& uniform;
};
/**
The class random_generator can be used in the STL generate function
to easily generate random ints between [0, _max).
*/
template <class T = uint32> class random_generator
{
public :
random_generator(int _max, eoRng& _rng = rng) : maxim(_max), random(_rng) {}
virtual T operator()(void) { return (T) random.random(max); }
private :
T maxim;
eoRng& random;
};
/**
The class normal_generator can be used in the STL generate function
to easily generate gaussian distributed floats and doubles. The user
can supply a standard deviation which defaults to 1.
*/
template <class T = double> class normal_generator
{
public :
normal_generator(T _stdev = T(1.0), eoRng& _rng = rng) : stdev(_stdev), normal(_rng) {}
virtual T operator()(void) { return (T) normal.normal(stdev); }
private :
T stdev;
eoRng& normal;
};
// Implementation of some eoRng members.... Don't mind the mess, it does work.
#define hiBit(u) ((u) & 0x80000000U) // mask all but highest bit of u
#define loBit(u) ((u) & 0x00000001U) // mask all but lowest bit of u
#define loBits(u) ((u) & 0x7FFFFFFFU) // mask the highest bit of u
#define mixBits(u, v) (hiBit(u)|loBits(v)) // move hi bit of u to hi bit of v
inline void eoRng::initialize(uint32 seed)
{
//
// We initialize state[0..(N-1)] via the generator
//
// x_new = (69069 * x_old) mod 2^32
//
// from Line 15 of Table 1, p. 106, Sec. 3.3.4 of Knuth's
// _The Art of Computer Programming_, Volume 2, 3rd ed.
//
// Notes (SJC): I do not know what the initial state requirements
// of the Mersenne Twister are, but it seems this seeding generator
// could be better. It achieves the maximum period for its modulus
// (2^30) iff x_initial is odd (p. 20-21, Sec. 3.2.1.2, Knuth); if
// x_initial can be even, you have sequences like 0, 0, 0, ...;
// 2^31, 2^31, 2^31, ...; 2^30, 2^30, 2^30, ...; 2^29, 2^29 + 2^31,
// 2^29, 2^29 + 2^31, ..., etc. so I force seed to be odd below.
//
// Even if x_initial is odd, if x_initial is 1 mod 4 then
//
// the lowest bit of x is always 1,
// the next-to-lowest bit of x is always 0,
// the 2nd-from-lowest bit of x alternates ... 0 1 0 1 0 1 0 1 ... ,
// the 3rd-from-lowest bit of x 4-cycles ... 0 1 1 0 0 1 1 0 ... ,
// the 4th-from-lowest bit of x has the 8-cycle ... 0 0 0 1 1 1 1 0 ... ,
// ...
//
// and if x_initial is 3 mod 4 then
//
// the lowest bit of x is always 1,
// the next-to-lowest bit of x is always 1,
// the 2nd-from-lowest bit of x alternates ... 0 1 0 1 0 1 0 1 ... ,
// the 3rd-from-lowest bit of x 4-cycles ... 0 0 1 1 0 0 1 1 ... ,
// the 4th-from-lowest bit of x has the 8-cycle ... 0 0 1 1 1 1 0 0 ... ,
// ...
//
// The generator's potency (min. s>=0 with (69069-1)^s = 0 mod 2^32) is
// 16, which seems to be alright by p. 25, Sec. 3.2.1.3 of Knuth. It
// also does well in the dimension 2..5 spectral tests, but it could be
// better in dimension 6 (Line 15, Table 1, p. 106, Sec. 3.3.4, Knuth).
//
// Note that the random number user does not see the values generated
// here directly since restart() will always munge them first, so maybe
// none of all of this matters. In fact, the seed values made here could
// even be extra-special desirable if the Mersenne Twister theory says
// so-- that's why the only change I made is to restrict to odd seeds.
//
left = -1;
register uint32 x = (seed | 1U) & 0xFFFFFFFFU, *s = state;
register int j;
for(left=0, *s++=x, j=N; --j;
*s++ = (x*=69069U) & 0xFFFFFFFFU);
}
inline uint32 eoRng::restart(void)
{
register uint32 *p0=state, *p2=state+2, *pM=state+M, s0, s1;
register int j;
left=N-1, next=state+1;
for(s0=state[0], s1=state[1], j=N-M+1; --j; s0=s1, s1=*p2++)
*p0++ = *pM++ ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
for(pM=state, j=M; --j; s0=s1, s1=*p2++)
*p0++ = *pM++ ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
s1=state[0], *p0 = *pM ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
s1 ^= (s1 >> 11);
s1 ^= (s1 << 7) & 0x9D2C5680U;
s1 ^= (s1 << 15) & 0xEFC60000U;
return(s1 ^ (s1 >> 18));
}
inline uint32 eoRng::rand(void)
{
uint32 y;
if(--left < 0)
return(restart());
y = *next++;
y ^= (y >> 11);
y ^= (y << 7) & 0x9D2C5680U;
y ^= (y << 15) & 0xEFC60000U;
return(y ^ (y >> 18));
}
inline double eoRng::normal(void)
{
if (cached)
{
cached = false;
return cacheValue;
}
float rSquare, factor, var1, var2;
do
{
var1 = 2.0 * uniform() - 1.0;
var2 = 2.0 * uniform() - 1.0;
rSquare = var1 * var1 + var2 * var2;
}
while (rSquare >= 1.0 || rSquare == 0.0);
factor = sqrt(-2.0 * log(rSquare) / rSquare);
cacheValue = var1 * factor;
cached = true;
return (var2 * factor);
}
#endif
/*
* Random number generator adapted from (see comments below)
*
* The random number generator is modified into a class
* by Maarten Keijzer (mak@dhi.dk). Also added the Box-Muller
* transformation to generate normal deviates.
*
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
/* ************ DOCUMENTATION IN ORIGINAL FILE *********************/
// This is the ``Mersenne Twister'' random number generator MT19937, which
// generates pseudorandom integers uniformly distributed in 0..(2^32 - 1)
// starting from any odd seed in 0..(2^32 - 1). This version is a recode
// by Shawn Cokus (Cokus@math.washington.edu) on March 8, 1998 of a version by
// Takuji Nishimura (who had suggestions from Topher Cooper and Marc Rieffel in
// July-August 1997).
//
// Effectiveness of the recoding (on Goedel2.math.washington.edu, a DEC Alpha
// running OSF/1) using GCC -O3 as a compiler: before recoding: 51.6 sec. to
// generate 300 million random numbers; after recoding: 24.0 sec. for the same
// (i.e., 46.5% of original time), so speed is now about 12.5 million random
// number generations per second on this machine.
//
// According to the URL <http://www.math.keio.ac.jp/~matumoto/emt.html>
// (and paraphrasing a bit in places), the Mersenne Twister is ``designed
// with consideration of the flaws of various existing generators,'' has
// a period of 2^19937 - 1, gives a sequence that is 623-dimensionally
// equidistributed, and ``has passed many stringent tests, including the
// die-hard test of G. Marsaglia and the load test of P. Hellekalek and
// S. Wegenkittl.'' It is efficient in memory usage (typically using 2506
// to 5012 bytes of static data, depending on data type sizes, and the code
// is quite short as well). It generates random numbers in batches of 624
// at a time, so the caching and pipelining of modern systems is exploited.
// It is also divide- and mod-free.
//
// This library is free software; you can redistribute it and/or modify it
// under the terms of the GNU Library General Public License as published by
// the Free Software Foundation (either version 2 of the License or, at your
// option, any later version). This library 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 Library General Public License for more details. You should have
// received a copy of the GNU Library General Public License along with this
// library; if not, write to the Free Software Foundation, Inc., 59 Temple
// Place, Suite 330, Boston, MA 02111-1307, USA.
//
// The code as Shawn received it included the following notice:
//
// Copyright (C) 1997 Makoto Matsumoto and Takuji Nishimura. When
// you use this, send an e-mail to <matumoto@math.keio.ac.jp> with
// an appropriate reference to your work.
//
// It would be nice to CC: <Cokus@math.washington.edu> when you write.
//
//
// uint32 must be an unsigned integer type capable of holding at least 32
// bits; exactly 32 should be fastest, but 64 is better on an Alpha with
// GCC at -O3 optimization so try your options and see what's best for you
//
/* ************ END DOCUMENTATION IN ORIGINAL FILE *********************/
#ifndef EO_RANDOM_NUMBER_GENERATOR
#define EO_RANDOM_NUMBER_GENERATOR
#include <ctime>
#include <eoPersistent.h>
#include <eoObject.h>
// TODO: check for various compilers if this is exactly 32 bits
// Unfortunately MSVC's preprocessor does not comprehends sizeof()
// so neat preprocessing tricks will not work
typedef unsigned long uint32; // Compiler and platform dependent!
//-----------------------------------------------------------------------------
// eoRng
//-----------------------------------------------------------------------------
/**
eoRng is a persitent class that uses the ``Mersenne Twister'' random number generator MT19937
for generating random numbers. The various member functions implement useful functions
for evolutionary algorithms. Included are: rand(), random(), flip() and normal().
Note for people porting EO to other platforms: please make sure that the typedef
uint32 in the file eoRng.h is exactly 32 bits long. It may be longer, but not
shorter. If it is longer, file compatibility between EO on different platforms
may be broken.
*/
class eoRng : public eoObject, public eoPersistent
{
public :
/**
ctor takes a random seed; if you want another seed, use reseed
@see reseed
*/
eoRng(uint32 s = (uint32) time(0) ) : state(0), next(0), left(-1), cached(false), N(624), M(397), K(0x9908B0DFU) {
state = new uint32[N+1];
initialize(s);
}
~eoRng(void)
{
delete [] state;
}
/**
Re-initializes the Random Number Generator.
*/
void reseed(uint32 s)
{
initialize(s);
}
/**
uniform(m = 1.0) returns a random double in the range [0, m)
*/
double uniform(double m = 1.0)
{ // random number between [0, m]
return m * double(rand()) / double(rand_max());
}
/**
random() returns a random integer in the range [0, m)
*/
uint32 random(uint32 m)
{
return uint32(uniform() * double(m));
}
/**
flip() tosses a biased coin such that flip(x/100.0) will
returns true x% of the time
*/
bool flip(float bias)
{
return uniform() < bias;
}
/**
normal() zero mean gaussian deviate with standard deviation of 1
*/
double normal(void); // gaussian mutation, stdev 1
/**
normal(stdev) zero mean gaussian deviate with user defined standard deviation
*/
double normal(double stdev)
{
return stdev * normal();
}
/**
normal(mean, stdev) user defined mean gaussian deviate with user defined standard deviation
*/
double normal(double mean, double stdev)
{
return mean + normal(stdev);
}
/**
rand() returns a random number in the range [0, rand_max)
*/
uint32 rand();
/**
rand_max() the maximum returned by rand()
*/
uint32 rand_max(void) const { return (uint32) 0xffffffff; }
/**
roulette_wheel(vec, total = 0) does a roulette wheel selection
on the input vector vec. If the total is not supplied, it is
calculated. It returns an integer denoting the selected argument.
*/
template <class T>
int roulette_wheel(const std::vector<T>& vec, T total = 0)
{
if (total == 0)
{ // count
for (int i = 0; i < vec.size(); ++i)
total += vec[i];
}
float change = uniform() * total;
int i = 0;
while (change > 0)
{
change -= vec[i++];
}
return --i;
}
///
void printOn(ostream& _os) const
{
for (int i = 0; i < N; ++i)
{
_os << state[i] << ' ';
}
_os << int(next - state) << ' ';
_os << left << ' ' << cached << ' ' << cacheValue;
}
///
void readFrom(istream& _is)
{
for (int i = 0; i < N; ++i)
{
_is >> state[i];
}
int n;
_is >> n;
next = state + n;
_is >> left;
_is >> cached;
_is >> cacheValue;
}
private :
uint32 restart(void);
void initialize(uint32 seed);
uint32* state; // the array for the state
uint32* next;
int left;
bool cached;
float cacheValue;
const int N;
const int M;
const uint32 K; // a magic constant
/**
Private copy ctor and assignment operator to make sure that
nobody accidentally copies the random number generator.
If you want similar RNG's, make two RNG's and initialize
them with the same seed.
*/
eoRng (const eoRng&); // no implementation
eoRng& operator=(const eoRng&); // dito
};
/**
The one and only global eoRng object
*/
static eoRng rng;
/**
The class uniform_generator can be used in the STL generate function
to easily generate random floats and doubles between [0, _max). _max
defaults to 1.0
*/
template <class T = double> class uniform_generator
{
public :
uniform_generator(T _max = T(1.0), eoRng& _rng = rng) : maxim(_max), uniform(_rng) {}
virtual T operator()(void) { return (T) uniform.uniform(maxim); }
private :
T maxim;
eoRng& uniform;
};
/**
The class random_generator can be used in the STL generate function
to easily generate random ints between [0, _max).
*/
template <class T = uint32> class random_generator
{
public :
random_generator(int _max, eoRng& _rng = rng) : maxim(_max), random(_rng) {}
virtual T operator()(void) { return (T) random.random(max); }
private :
T maxim;
eoRng& random;
};
/**
The class normal_generator can be used in the STL generate function
to easily generate gaussian distributed floats and doubles. The user
can supply a standard deviation which defaults to 1.
*/
template <class T = double> class normal_generator
{
public :
normal_generator(T _stdev = T(1.0), eoRng& _rng = rng) : stdev(_stdev), normal(_rng) {}
virtual T operator()(void) { return (T) normal.normal(stdev); }
private :
T stdev;
eoRng& normal;
};
// Implementation of some eoRng members.... Don't mind the mess, it does work.
#define hiBit(u) ((u) & 0x80000000U) // mask all but highest bit of u
#define loBit(u) ((u) & 0x00000001U) // mask all but lowest bit of u
#define loBits(u) ((u) & 0x7FFFFFFFU) // mask the highest bit of u
#define mixBits(u, v) (hiBit(u)|loBits(v)) // move hi bit of u to hi bit of v
inline void eoRng::initialize(uint32 seed)
{
//
// We initialize state[0..(N-1)] via the generator
//
// x_new = (69069 * x_old) mod 2^32
//
// from Line 15 of Table 1, p. 106, Sec. 3.3.4 of Knuth's
// _The Art of Computer Programming_, Volume 2, 3rd ed.
//
// Notes (SJC): I do not know what the initial state requirements
// of the Mersenne Twister are, but it seems this seeding generator
// could be better. It achieves the maximum period for its modulus
// (2^30) iff x_initial is odd (p. 20-21, Sec. 3.2.1.2, Knuth); if
// x_initial can be even, you have sequences like 0, 0, 0, ...;
// 2^31, 2^31, 2^31, ...; 2^30, 2^30, 2^30, ...; 2^29, 2^29 + 2^31,
// 2^29, 2^29 + 2^31, ..., etc. so I force seed to be odd below.
//
// Even if x_initial is odd, if x_initial is 1 mod 4 then
//
// the lowest bit of x is always 1,
// the next-to-lowest bit of x is always 0,
// the 2nd-from-lowest bit of x alternates ... 0 1 0 1 0 1 0 1 ... ,
// the 3rd-from-lowest bit of x 4-cycles ... 0 1 1 0 0 1 1 0 ... ,
// the 4th-from-lowest bit of x has the 8-cycle ... 0 0 0 1 1 1 1 0 ... ,
// ...
//
// and if x_initial is 3 mod 4 then
//
// the lowest bit of x is always 1,
// the next-to-lowest bit of x is always 1,
// the 2nd-from-lowest bit of x alternates ... 0 1 0 1 0 1 0 1 ... ,
// the 3rd-from-lowest bit of x 4-cycles ... 0 0 1 1 0 0 1 1 ... ,
// the 4th-from-lowest bit of x has the 8-cycle ... 0 0 1 1 1 1 0 0 ... ,
// ...
//
// The generator's potency (min. s>=0 with (69069-1)^s = 0 mod 2^32) is
// 16, which seems to be alright by p. 25, Sec. 3.2.1.3 of Knuth. It
// also does well in the dimension 2..5 spectral tests, but it could be
// better in dimension 6 (Line 15, Table 1, p. 106, Sec. 3.3.4, Knuth).
//
// Note that the random number user does not see the values generated
// here directly since restart() will always munge them first, so maybe
// none of all of this matters. In fact, the seed values made here could
// even be extra-special desirable if the Mersenne Twister theory says
// so-- that's why the only change I made is to restrict to odd seeds.
//
left = -1;
register uint32 x = (seed | 1U) & 0xFFFFFFFFU, *s = state;
register int j;
for(left=0, *s++=x, j=N; --j;
*s++ = (x*=69069U) & 0xFFFFFFFFU);
}
inline uint32 eoRng::restart(void)
{
register uint32 *p0=state, *p2=state+2, *pM=state+M, s0, s1;
register int j;
left=N-1, next=state+1;
for(s0=state[0], s1=state[1], j=N-M+1; --j; s0=s1, s1=*p2++)
*p0++ = *pM++ ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
for(pM=state, j=M; --j; s0=s1, s1=*p2++)
*p0++ = *pM++ ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
s1=state[0], *p0 = *pM ^ (mixBits(s0, s1) >> 1) ^ (loBit(s1) ? K : 0U);
s1 ^= (s1 >> 11);
s1 ^= (s1 << 7) & 0x9D2C5680U;
s1 ^= (s1 << 15) & 0xEFC60000U;
return(s1 ^ (s1 >> 18));
}
inline uint32 eoRng::rand(void)
{
uint32 y;
if(--left < 0)
return(restart());
y = *next++;
y ^= (y >> 11);
y ^= (y << 7) & 0x9D2C5680U;
y ^= (y << 15) & 0xEFC60000U;
return(y ^ (y >> 18));
}
inline double eoRng::normal(void)
{
if (cached)
{
cached = false;
return cacheValue;
}
float rSquare, factor, var1, var2;
do
{
var1 = 2.0 * uniform() - 1.0;
var2 = 2.0 * uniform() - 1.0;
rSquare = var1 * var1 + var2 * var2;
}
while (rSquare >= 1.0 || rSquare == 0.0);
factor = sqrt(-2.0 * log(rSquare) / rSquare);
cacheValue = var1 * factor;
cached = true;
return (var2 * factor);
}
#endif

View file

@ -0,0 +1,51 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoRandomIndiSelector.h
Selects individuals at random.
(c) Maarten Keijzer (mkeijzer@mad.scientist.com) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoRandomIndiSelector_h
#define eoRandomIndiSelector_h
#include "eoIndiSelector.h"
/**
* eoRandomSelector: just selects a random child
*/
template <class EOT>
class eoRandomIndiSelector : public eoPopIndiSelector<EOT>
{
public :
eoRandomIndiSelector(void) : eoPopIndiSelector<EOT>() {}
virtual ~eoRandomIndiSelector(void) {}
/// very complex function that returns just an individual
const EOT& do_select(void)
{
return operator[](rng.random(size()));
}
};
#endif

View file

@ -0,0 +1,61 @@
/** -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoSequentialGOpSelector.h
Sequential Generalized Operator Selector.
(c) Maarten Keijzer (mkeijzer@mad.scientist.com), GeNeura Team 1998, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoSequentialGOpSelector_h
#define eoSequentialGOpSelector_h
//-----------------------------------------------------------------------------
#include "eoGOpSelector.h"
/** eoSequentialGOpSel: do proportional selection, but return a sequence of
operations to be applied one after the other.
*/
template <class EOT>
class eoSequentialGOpSel : public eoGOpSelector<EOT>
{
public :
virtual eoGeneralOp<EOT>& selectOp()
{
combined.clear();
for (int i = 0; i < size(); ++i)
{
if (operator[](i) == 0)
continue;
if (rng.flip(getRates()[i]))
combined.addOp(operator[](i));
}
return combined;
}
private :
eoCombinedOp<EOT> combined;
};
#endif

85
eo/src/eoSteadyStateEA.h Normal file
View file

@ -0,0 +1,85 @@
// -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
//-----------------------------------------------------------------------------
// eoSteadyStateEA.h
// (c) GeNeura Team, 2000
/*
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
//-----------------------------------------------------------------------------
#ifndef _eoSteadyStateEA_h
#define _eoSteadyStateEA_h
//-----------------------------------------------------------------------------
#include "eoSteadyStateGeneration.h" // eoPop
#include <eoTerm.h>
/** EOSteadyStateEA:
An easy-to-use evolutionary algorithm, just supply
a general operator selector, a selector for choosing the ones
to reproduce and an eoSteadyStateInserter that takes care of evaluating
and inserter the guy/girl in the steady state population.
*/
template<class EOT> class eoSteadyStateEA: public eoAlgo<EOT>
{
public:
/// Constructor.
eoSteadyStateEA(
eoGOpSelector<EOT>& _opSelector,
eoPopIndiSelector<EOT>& _selector,
eoSteadyStateInserter<EOT>& _inserter,
eoTerm<Chrom>& _terminator,
unsigned _steps = 0 )
: step(_opSelector, _selector, _inserter),
terminator( _terminator)
{};
/// Constructor from an already created generation
eoSteadyStateEA(eoSteadyStateGeneration<EOT>& _gen,
eoTerm<EOT>& _terminator):
step(_gen),
terminator( _terminator){};
/// Apply one generation of evolution to the population.
virtual void operator()(eoPop<Chrom>& pop) {
do {
try
{
step(pop);
}
catch (exception& e)
{
string s = e.what();
s.append( " in eoSteadyStateEA ");
throw runtime_error( s );
}
} while ( terminator( pop ) );
}
/// Class name.
string className() const { return "eoSteadyStateEA"; }
private:
eoSteadyStateGeneration<EOT> step;
eoTerm<EOT>& terminator;
};
//-----------------------------------------------------------------------------
#endif eoEasyEA_h

View file

@ -0,0 +1,88 @@
// -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
//-----------------------------------------------------------------------------
// eoSteadyStateGeneration.h
// (c) GeNeura Team, 1998
/*
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
//-----------------------------------------------------------------------------
#ifndef eoSteadyStateGeneration_h
#define eoSteadyStateGeneration_h
//-----------------------------------------------------------------------------
#include <eoAlgo.h> // eoPop
#include <eoEvalFunc.h>
#include <eoPopOps.h> // eoSelect, eoTranform, eoMerge
#include "eoGOpSelector.h"
#include "eoIndiSelector.h"
#include "eoSteadyStateInserter.h"
//-----------------------------------------------------------------------------
/** eoSteadyStateGeneration
* Single step of a steady state evolutionary algorithm.
* Proceeds by updating one individual at a time, by first selecting parents,
* creating one or more children and subsequently overwrite (a) bad individual(s)
*/
template<class EOT> class eoSteadyStateGeneration: public eoAlgo<EOT>
{
public:
/// Constructor.
eoSteadyStateGeneration(
eoGOpSelector<EOT>& _opSelector,
eoPopIndiSelector<EOT>& _selector,
eoSteadyStateInserter<EOT>& _inserter,
unsigned _steps = 0) :
opSelector(_opSelector),
selector(_selector),
inserter(_inserter) ,
steps(_steps) {};
/// Apply one generation of evolution to the population.
virtual void operator()(eoPop<EOT>& pop)
{
unsigned nSteps = steps;
if (nSteps == 0)
{
nSteps = pop.size(); // make a 'generation equivalent'
}
for (unsigned i = 0; i < nSteps; ++i)
{
opSelector.selectOp()(selector(pop), inserter(pop));
}
}
/// Class name.
string className() const { return "eoSteadyStateGeneration"; }
private:
eoGOpSelector<EOT>& opSelector;
eoPopIndiSelector<EOT>& selector;
eoSteadyStateInserter<EOT>& inserter;
unsigned steps;
};
//-----------------------------------------------------------------------------
#endif eoGeneration_h

View file

@ -0,0 +1,51 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoSteadyStateInserter.h
Still abstract population insertion operator that is initialized with
and eoEvalFunc object to be able to evaluate individuals before inserting
them.
(c) Maarten Keijzer (mak@dhi.dk) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoSteadyStateInserter_h
#define eoSteadyStateInserter_h
#include "eoEvalFunc.h"
/**
* eoSteadyStateInserter: Interface class that enables an operator to update
* a population with a new individual... it contains an eoEvalFunc object to
* make sure that every individual is evaluated before it is inserted
*/
template <class EOT>
class eoSteadyStateInserter : public eoPopInserter<EOT>
{
public :
eoSteadyStateInserter(eoEvalFunc<EOT>& _eval) : eval(_eval) , eoPopInserter<EOT>() {}
protected :
eoEvalFunc<EOT>& eval;
};
#endif

View file

@ -0,0 +1,73 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoStochTournamentInserter.h
Concrete steady state inserter. It is initialized with a population and
inserts individuals in the population based on an inverse stochastic
tournament
(c) Maarten Keijzer (mkeijzer@mad.scientist.com) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef eoStochTournamentInserter_h
#define eoStochTournamentInserter_h
#include "eoSteadyStateInserter.h"
#include "selectors.h"
/**
* eoDetTournamentInserter: Uses an inverse stochastic tournament to figure
* out who gets overridden by the new individual. It resets the fitness of the
* individual.
*/
template <class EOT>
class eoStochTournamentInserter : public eoSteadyStateInserter<EOT>
{
public :
eoStochTournamentInserter(eoEvalFunc<EOT>& _eval, double _t_rate) : t_rate(_t_rate), eoSteadyStateInserter<EOT>(_eval)
{
if (t_rate < 0.5)
{ // warning, error?
t_rate = 0.55;
}
if (t_rate >= 1.0)
{
t_rate = 0.99; // 1.0 would mean deterministic tournament
}
}
void insert(const EOT& _eot)
{
EOT& eo = inverse_stochastic_tournament<EOT>(pop(), t_rate);
eo = _eot; // overwrite loser of tournament
eo.invalidate();
eval(eo); // Evaluate after insert
}
string className(void) const { return "eoStochTournamentInserter"; }
private :
double t_rate;
};
#endif

221
eo/src/eoWrappedOps.h Normal file
View file

@ -0,0 +1,221 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
eoWrappedOps.h
Derived from the General genetic operator, which can be used to wrap any unary or binary
operator. File also contains the eoCombinedOp, needed by the eoSequentialGOpSelector
(c) Maarten Keijzer (mak@dhi.dk) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
//-----------------------------------------------------------------------------
#ifndef eoWrappedOps_h
#define eoWrappedOps_h
//-----------------------------------------------------------------------------
#include <eoOp.h> // eoOp, eoMonOp, eoBinOp
#include "eoRNG.h"
using namespace std;
/// Wraps monary operators
template <class EOT>
class eoWrappedMonOp : public eoGeneralOp<EOT>
{
public :
///
eoWrappedMonOp(const eoMonOp<EOT>& _op) : eoGeneralOp<EOT>(), op(_op) {};
///
virtual ~eoWrappedMonOp() {}
/// Instantiates the abstract method
void operator()( eoIndiSelector<EOT>& _in,
eoInserter<EOT>& _out) const {
EOT result = _in.select();
op( result );
_out.insert(result);
}
///
virtual string className() const {return "eoWrappedMonOp";};
private :
const eoMonOp<EOT>& op;
};
/// Wraps binary operators
template <class EOT>
class eoWrappedBinOp : public eoGeneralOp<EOT>
{
public :
///
eoWrappedBinOp(const eoBinOp<EOT>& _op) : eoGeneralOp<EOT>(), op(_op) {}
///
virtual ~eoWrappedBinOp() {}
/// Instantiates the abstract method. EOT should have copy ctor.
void operator()(eoIndiSelector<EOT>& _in,
eoInserter<EOT>& _out) const {
EOT out1 = _in.select();
const EOT& out2 = _in.select();
op(out1, out2);
_out.insert(out1);
}
///
virtual string className() const {return "eoWrappedBinOp";};
private :
const eoBinOp<EOT>& op;
};
/// Wraps Quadratic operators
template <class EOT>
class eoWrappedQuadraticOp : public eoGeneralOp<EOT>
{
public :
///
eoWrappedQuadraticOp(const eoQuadraticOp<EOT>& _op) : eoGeneralOp<EOT>(), op(_op) {}
///
virtual ~eoWrappedQuadraticOp() {}
/// Instantiates the abstract method. EOT should have copy ctor.
void operator()(eoIndiSelector<EOT>& _in,
eoInserter<EOT>& _out) const {
EOT out1 = _in.select();
EOT out2 = _in.select();
op(out1, out2);
_out.insert(out1);
_out.insert(out2);
}
///
virtual string className() const {return "eoWrappedQuadraticOp";};
private :
const eoQuadraticOp<EOT>& op;
};
/// Combines several ops
template <class EOT>
class eoCombinedOp : public eoGeneralOp<EOT>
{
public :
///
eoCombinedOp() : eoGeneralOp<EOT>() {}
///
virtual ~eoCombinedOp() {}
/// Adds a new operator to the combined Op
void addOp(eoGeneralOp<EOT>* _op)
{
ops.push_back(_op);
}
/// Erases all operators added so far
void clear(void) {
ops.resize(0);
}
/// Helper class to make sure that stuff that is inserted will be used again with the next operator
template <class EOT>
class eoIndiSelectorInserter : public eoIndiSelector<EOT>, public eoInserter<EOT>
{
public :
eoIndiSelectorInserter(eoIndiSelector<EOT>& _in)
: eoIndiSelector<EOT>(), eoInserter<EOT>(), in(_in)
{}
size_t size() const { return in.size(); }
const EOT& operator[](size_t _n) const { return in[_n]; }
const EOT& select(void)
{
if (results.empty())
{
return in.select();
}
// else we use the previously inserted individual,
// an iterator to it is stored in 'results', but the memory
// is kept by 'intermediate'.
list<EOT>::iterator it = *results.begin();
results.pop_front();
return *it;
}
void insert(const EOT& _eot)
{
intermediate.push_front(_eot);
results.push_front(intermediate.begin());
}
void fill(eoInserter<EOT>& _out)
{
typedef list<list<EOT>::iterator>::iterator Iterator;
for (Iterator it = results.begin(); it != results.end(); ++it)
{
_out.insert(**it);
}
results.clear();
intermediate.clear(); // reclaim memory
}
private :
eoIndiSelector<EOT>& in;
// using lists as we need to push and pop a lot
// 'results' are iterators to the contents of 'intermediate'
// to prevent copying to and from intermediate...
list<list<EOT>::iterator> results;
list<EOT> intermediate;
};
/// Applies all ops in the combined op
void operator()( eoIndiSelector<EOT>& _in,
eoInserter<EOT>& _out ) const {
eoIndiSelectorInserter<EOT> in_out(_in);
for (size_t i = 0; i < ops.size(); ++i)
{
(*ops[i])(in_out, in_out);
}
in_out.fill(_out);
}
private :
vector<eoGeneralOp<EOT>* > ops;
};
#endif eoGeneral_h

85
eo/src/rnd_generators.h Normal file
View file

@ -0,0 +1,85 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
rnd_generators.h
Some utility functors for generating random generators:
uniform_generator : generates uniform floats or doubles
random_generator : generates unsigneds, ints etc.
normal_generator : normally distributed floats or doubles
(c) Maarten Keijzer (mak@dhi.dk) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
//-----------------------------------------------------------------------------
#ifndef eoRND_GENERATORS_H
#define eoRND_GENERATORS_H
#include "eoRNG.h"
/**
The class uniform_generator can be used in the STL generate function
to easily generate random floats and doubles between [0, _max). _max
defaults to 1.0
*/
template <class T = double> class uniform_generator
{
public :
uniform_generator(T _max = T(1.0), eoRng& _rng = rng) : maxim(_max), uniform(_rng) {}
virtual T operator()(void) { return (T) uniform.uniform(maxim); }
private :
T maxim;
eoRng& uniform;
};
/**
The class random_generator can be used in the STL generate function
to easily generate random ints between [0, _max).
*/
template <class T = uint32> class random_generator
{
public :
random_generator(int _max, eoRng& _rng = rng) : maxim(_max), random(_rng) {}
virtual T operator()(void) { return (T) random.random(max); }
private :
T maxim;
eoRng& random;
};
/**
The class normal_generator can be used in the STL generate function
to easily generate gaussian distributed floats and doubles. The user
can supply a standard deviation which defaults to 1.
*/
template <class T = double> class normal_generator
{
public :
normal_generator(T _stdev = T(1.0), eoRng& _rng = rng) : stdev(_stdev), normal(_rng) {}
virtual T operator()(void) { return (T) normal.normal(stdev); }
private :
T stdev;
eoRng& normal;
};
#endif

313
eo/src/selectors.h Normal file
View file

@ -0,0 +1,313 @@
/* -*- mode: c++; c-indent-level: 4; c++-member-init-indent: 8; comment-column: 35; -*-
-----------------------------------------------------------------------------
selectors.h
A bunch of useful selector functions. They generally have three forms:
template <class It>
It select(It begin, It end, params, eoRng& gen = rng);
template <class EOT>
const EOT& select(const eoPop<EOT>& pop, params, eoRng& gen = rng);
template <class EOT>
EOT& select(eoPop<EOT>& pop, params, eoRng& gen = rng);
where select is one of: roulette_wheel, deterministic_tournament
and stochastic_tournament (at the moment).
(c) Maarten Keijzer (mak@dhi.dk) and GeNeura Team, 1999, 2000
This library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2 of the License, or (at your option) any later version.
This library 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
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with this library; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
Contact: todos@geneura.ugr.es, http://geneura.ugr.es
*/
#ifndef SELECT__H
#define SELECT__H
#include "eoRNG.h"
#include "eoException.h"
template <class EOT>
bool minimizing_fitness()
{
EOT eo1; // Assuming people don't do anything fancy in the default constructor!
EOT eo2;
/* Dear user, when the two line below do not compile you are most
likely not working with scalar fitness values. In that case we're sorry
but you cannot use lottery or roulette_wheel selection...
*/
eo1.fitness(0.0); // tried to cast it to an EOT::Fitness, but for some reason GNU barfs on this
eo2.fitness(1.0);
return eo2 < eo1; // check whether we have a minimizing fitness
};
inline double scale_fitness(const std::pair<double, double>& _minmax, double _value)
{
if (_minmax.first == _minmax.second)
{
return 0.0; // no differences in fitness, population converged!
}
// else
return (_value - _minmax.first) / (_minmax.second - _minmax.first);
}
template <class It>
double sum_fitness(It begin, It end)
{
double sum = 0.0;
for (; begin != end; ++begin)
{
double v = static_cast<double>(begin->fitness());
if (v < 0.0)
throw eoNegativeFitnessException();
sum += v;
}
return sum;
}
template <class EOT>
double sum_fitness(const eoPop<EOT>& _pop)
{
return sum_fitness(_pop.begin(), _pop.end());
}
template <class EOT>
double sum_fitness(const eoPop<EOT>& _pop, std::pair<double, double>& _minmax)
{
eoPop<EOT>::const_iterator it = _pop.begin();
_minmax.first = it->fitness();
_minmax.second = it++->fitness();
for(; it != _pop.end(); ++it)
{
double v = static_cast<double>(it->fitness());
_minmax.first = std::min(_minmax.first, v);
_minmax.second = std::max(_minmax.second, v);
rawTotal += v;
}
if (minimizing_fitness<EOT>())
{
std::swap(_minmax.first, _minmax.second);
}
scaledTotal = 0.0;
// unfortunately a second loop is neccessary to scale the fitness
for (it = _pop.begin(); it != _pop.end(); ++it)
{
double v = scale_fitness(static_cast<double>(it->fitness()));
scaledTotal += v;
}
}
template <class It>
It roulette_wheel(It _begin, It _end, double total, eoRng& _gen = rng)
{
float roulette = _gen.uniform(total);
It i = _begin;
while (roulette > 0.0)
{
roulette -= static_cast<double>(*(i++));
}
return --i;
}
template <class EOT>
const EOT& roulette_wheel(const eoPop<EOT>& _pop, double total, eoRng& _gen = rng)
{
float roulette = _gen.uniform(total);
eoPop<EOT>::const_iterator i = _pop.begin();
while (roulette > 0.0)
{
roulette -= static_cast<double>((i++)->fitness());
}
return *--i;
}
template <class EOT>
EOT& roulette_wheel(eoPop<EOT>& _pop, double total, eoRng& _gen = rng)
{
float roulette = _gen.uniform(total);
eoPop<EOT>::iterator i = _pop.begin();
while (roulette > 0.0)
{
roulette -= static_cast<double>((i++)->fitness());
}
return *--i;
}
template <class It>
It deterministic_tournament(It _begin, It _end, unsigned _t_size, eoRng& _gen = rng)
{
It best = _begin + _gen.random(_end - _begin);
for (unsigned i = 0; i < _t_size - 1; ++i)
{
It competitor = _begin + _gen.random(_end - _begin);
if (*best < *competitor)
{
best = competitor;
}
}
return best;
}
template <class EOT>
const EOT& deterministic_tournament(const eoPop<EOT>& _pop, unsigned _t_size, eoRng& _gen = rng)
{
return *deterministic_tournament(_pop.begin(), _pop.end(), _t_size, _gen);
}
template <class EOT>
EOT& deterministic_tournament(eoPop<EOT>& _pop, unsigned _t_size, eoRng& _gen = rng)
{
return *deterministic_tournament(_pop.begin(), _pop.end(), _t_size, _gen);
}
template <class It>
It inverse_deterministic_tournament(It _begin, It _end, unsigned _t_size, eoRng& _gen = rng)
{
It worst = _begin + _gen.random(_end - _begin);
for (unsigned i = 0; i < _t_size - 1; ++i)
{
It competitor = _begin + _gen.random(_end - _begin);
if (competitor == worst)
{
--i;
continue; // try again
}
if (*competitor < *worst)
{
worst = competitor;
}
}
return worst;
}
template <class EOT>
const EOT& inverse_deterministic_tournament(const eoPop<EOT>& _pop, unsigned _t_size, eoRng& _gen = rng)
{
return *inverse_deterministic_tournament<EOT>(_pop.begin(), _pop.end(), _t_size, _gen);
}
template <class EOT>
EOT& inverse_deterministic_tournament(eoPop<EOT>& _pop, unsigned _t_size, eoRng& _gen = rng)
{
return *inverse_deterministic_tournament(_pop.begin(), _pop.end(), _t_size, _gen);
}
template <class It>
It stochastic_tournament(It _begin, It _end, double _t_rate, eoRng& _gen = rng)
{
It i1 = _begin + _gen.random(_end - _begin);
It i2 = _begin + _gen.random(_end - _begin);
bool return_better = _gen.flip(_t_rate);
if (*i1 < *i2)
{
if (return_better) return i2;
// else
return i1;
}
else
{
if (return_better) return i1;
// else
}
// else
return i2;
}
template <class EOT>
const EOT& stochastic_tournament(const eoPop<EOT>& _pop, double _t_rate, eoRng& _gen = rng)
{
return *stochastic_tournament(_pop.begin(), _pop.end(), _t_rate, _gen);
}
template <class EOT>
EOT& stochastic_tournament(eoPop<EOT>& _pop, double _t_rate, eoRng& _gen = rng)
{
return *stochastic_tournament(_pop.begin(), _pop.end(), _t_rate, _gen);
}
template <class It>
It inverse_stochastic_tournament(It _begin, It _end, double _t_rate, eoRng& _gen = rng)
{
It i1 = _begin + _gen.random(_end - _begin);
It i2 = _begin + _gen.random(_end - _begin);
bool return_worse = _gen.flip(_t_rate);
if (*i1 < *i2)
{
if (return_worse) return i1;
// else
return i2;
}
else
{
if (return_worse) return i2;
// else
}
// else
return i1;
}
template <class EOT>
const EOT& inverse_stochastic_tournament(const eoPop<EOT>& _pop, double _t_rate, eoRng& _gen = rng)
{
return *inverse_stochastic_tournament(_pop.begin(), _pop.end(), _t_rate, _gen);
}
template <class EOT>
EOT& inverse_stochastic_tournament(eoPop<EOT>& _pop, double _t_rate, eoRng& _gen = rng)
{
return *inverse_stochastic_tournament(_pop.begin(), _pop.end(), _t_rate, _gen);
}
#endif