eda

eda − a C++ estimation of distribution library for orderings

#include <eda.h>

The eda C++ library, of which this file is a part, should be considered as being alpha testing software.

EDA 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 2 of the License, (here), or (at your option) any later version.

EDA 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 EDA; if not, write to the Free Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

Get a gzipped tar archive here.

eda is an estimation of distributions algorithm library for orderings on the numbers 0,1,...,N-1 for some positive N. Code that uses this library would look something like the following.

#include <eda.h>
#include <iostream>

#include <util.h>
using util::operator<<;
using util::operator>>;

class MyObjective : public Objective
{

vector<int> target;

public:

MyObjective(vector<int> &t) : Objective(maximize), target(t) {}
virtual ~MyObjective(){}
score_t raw(vector<int> &ind) {

int sum = 0;
for(vector<int>::iterator i = ind.begin() + 1; i != ind.end(); i++)

sum += static_cast<int>(target[*(i - 1)] <= target[*i]);

return static_cast<score_t>(sum);

}

};

int main()
{

vector<int> t;
std::cin >> t;
MyObjective o(t);
vector< pair<int,int> > zers; // disallowed transitions, here none
Distribution d(t.size(), zers);
EDA eda(d, o);
eda.run();
std::cout << eda.sol();

}

The above code comes from the file example.cc (here) supplied with this library, and can be compiled with (after installing the library and header files)
$ g++ -o example example.cc -leda -lgmpxx -lgmp
and run as
$ echo 3 2 1 4 100 5 | ./example
or can be compiled and run by
$ make test

The library consists of a number of files.

Header files

eda.h

Here is the main header file to be included by code that is to be linked with this library.

farith.h

Here is the file that is used by the library for floating point arithmetic. Uses the gmp library to avoid numerical instabilities.

objective.h

Here is the entry point to the library, defines the objective function.

parameter.h

Here is a sub-library that allows object-oriented access to command line parameters. Might be taken out of this library in the future.

compose2.h

Here is the file that is used by the command-line parameter part of the library

util.h

Here are utilities used by the library (templates)

CC files

eda.cc

farith.cc

parameter.cc

tsp(1)

Staal Vinterbo <staal@dsg.harvard.edu>

Copyright (C) 2005 Staal Vinterbo

This library is dependent on the GNU gmp library and has only been tested with g++ 3.3.{1,3}. The linking flags should include -lgmpxx -lgmp

Must be plenty. Please report them to the author. Thank you.