Growing up...

This commit is contained in:
blacklight 2009-08-16 11:09:42 +02:00
parent b62dfe3967
commit 25996a5e70
10 changed files with 169 additions and 72 deletions

View file

@ -1,5 +1,9 @@
--- Release 0.4 ---
2009-08-16 BlackLight <blacklight@autistici.org>
* neuron.cpp: Fixing propagate() function
2009-08-15 BlackLight <blacklight@autistici.org>
* Makefile: Now you compile Neural++ with -Wall -pedantic

View file

@ -0,0 +1,42 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html><head><meta http-equiv="Content-Type" content="text/html;charset=UTF-8">
<title>Neural++: Class Members</title>
<link href="doxygen.css" rel="stylesheet" type="text/css">
<link href="tabs.css" rel="stylesheet" type="text/css">
</head><body>
<!-- Generated by Doxygen 1.5.6 -->
<div class="navigation" id="top">
<div class="tabs">
<ul>
<li><a href="index.html"><span>Main&nbsp;Page</span></a></li>
<li class="current"><a href="namespaces.html"><span>Namespaces</span></a></li>
<li><a href="annotated.html"><span>Classes</span></a></li>
<li><a href="files.html"><span>Files</span></a></li>
</ul>
</div>
<div class="tabs">
<ul>
<li><a href="namespaces.html"><span>Namespace List</span></a></li>
<li class="current"><a href="namespacemembers.html"><span>Namespace&nbsp;Members</span></a></li>
</ul>
</div>
<div class="tabs">
<ul>
<li class="current"><a href="namespacemembers.html"><span>All</span></a></li>
<li><a href="namespacemembers_func.html"><span>Functions</span></a></li>
</ul>
</div>
</div>
<div class="contents">
Here is a list of all namespace members with links to the namespace documentation for each member:
<p>
<ul>
<li>df()
: <a class="el" href="namespaceneuralpp.html#43c8197cc83f65fa9676386579671aec">neuralpp</a>
</ul>
</div>
<hr size="1"><address style="text-align: right;"><small>Generated on Sat Aug 15 02:56:02 2009 for Neural++ by&nbsp;
<a href="http://www.doxygen.org/index.html">
<img src="doxygen.png" alt="doxygen" align="middle" border="0"></a> 1.5.6 </small></address>
</body>
</html>

View file

@ -0,0 +1,42 @@
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html><head><meta http-equiv="Content-Type" content="text/html;charset=UTF-8">
<title>Neural++: Class Members</title>
<link href="doxygen.css" rel="stylesheet" type="text/css">
<link href="tabs.css" rel="stylesheet" type="text/css">
</head><body>
<!-- Generated by Doxygen 1.5.6 -->
<div class="navigation" id="top">
<div class="tabs">
<ul>
<li><a href="index.html"><span>Main&nbsp;Page</span></a></li>
<li class="current"><a href="namespaces.html"><span>Namespaces</span></a></li>
<li><a href="annotated.html"><span>Classes</span></a></li>
<li><a href="files.html"><span>Files</span></a></li>
</ul>
</div>
<div class="tabs">
<ul>
<li><a href="namespaces.html"><span>Namespace List</span></a></li>
<li class="current"><a href="namespacemembers.html"><span>Namespace&nbsp;Members</span></a></li>
</ul>
</div>
<div class="tabs">
<ul>
<li><a href="namespacemembers.html"><span>All</span></a></li>
<li class="current"><a href="namespacemembers_func.html"><span>Functions</span></a></li>
</ul>
</div>
</div>
<div class="contents">
&nbsp;
<p>
<ul>
<li>df()
: <a class="el" href="namespaceneuralpp.html#43c8197cc83f65fa9676386579671aec">neuralpp</a>
</ul>
</div>
<hr size="1"><address style="text-align: right;"><small>Generated on Sat Aug 15 02:56:02 2009 for Neural++ by&nbsp;
<a href="http://www.doxygen.org/index.html">
<img src="doxygen.png" alt="doxygen" align="middle" border="0"></a> 1.5.6 </small></address>
</body>
</html>

View file

@ -7,6 +7,8 @@
#include <iostream>
#include <neural++.hpp>
using namespace std;
using namespace neuralpp;
int main() {

View file

@ -7,6 +7,8 @@
#include <iostream>
#include <neural++.hpp>
using namespace std;
using namespace neuralpp;
#define NETFILE "adder.net"

View file

@ -8,10 +8,12 @@
#include <iostream>
#include <neural++.hpp>
using namespace std;
using namespace neuralpp;
int main() {
NeuralNet net(2, 2, 1, 0.0005, 10000);
NeuralNet net(2, 2, 1, 0.005, 1000);
cout << "Training in progress - This may take a while...\n";
net.train("adder.xml", NeuralNet::file);

View file

@ -1,5 +1,5 @@
/**************************************************************************************************
* LibNeural++ v.0.2 - All-purpose library for managing neural networks *
* LibNeural++ v.0.4 - All-purpose library for managing neural networks *
* Copyright (C) 2009, BlackLight *
* *
* This program is free software: you can redistribute it and/or modify it under the terms of the *
@ -23,7 +23,6 @@
#include <cmath>
#include "neural++_exception.hpp"
using namespace std;
//! Default rand value: |sin(rand)|, always >= 0 and <= 1
#define RAND (double) ( (rand() / (RAND_MAX/2)) - 1)
@ -51,7 +50,8 @@ namespace neuralpp {
int epochs;
int ref_epochs;
double l_rate;
double ex;
//double ex;
std::vector<double> ex;
/**
* @brief It updates the weights of the net's synapsis through back-propagation.
@ -64,10 +64,10 @@ namespace neuralpp {
* In-class use only
* @param l Layer to commit the changes
*/
void commitChanges (Layer *l);
void commitChanges (Layer& l);
/**
* @brief Get the error made on the expected result as |v-v'|/v
* @brief Get the error made on the expected result as squared deviance
* @param ex Expected value
* @return Mean error
*/
@ -111,7 +111,7 @@ namespace neuralpp {
* @param file Binary file containing a neural network previously saved by save() method
* @throw NetworkFileNotFoundException
*/
NeuralNet (const string file) throw(NetworkFileNotFoundException);
NeuralNet (const std::string file) throw(NetworkFileNotFoundException);
/**
@ -139,21 +139,34 @@ namespace neuralpp {
* @brief It gets the output of the network in case the output layer contains more neurons
* @return A vector containing the output values of the network
*/
vector<double> getOutputs();
std::vector<double> getOutputs();
/**
* @brief It gets the value expected. Of course you should specify this when you
* @brief Get the expected value (in case you have an only neuron in output layer). Of course you should specify this when you
* build your network by using setExpected.
* @return The expected output value for a certain training phase
*/
double expected() const;
/**
* @brief It sets the value you expect from your network
* @brief Get the expected value (in case you have an only neuron in output layer). Of course you should specify this when you
* build your network by using setExpected.
* @return The expected output value for a certain training phase
*/
std::vector<double> getExpected() const;
/**
* @brief It sets the value you expect from your network (in case the network has an only neuron in its output layer)
* @param ex Expected output value
*/
void setExpected(double ex);
/**
* @brief Set the values you expect from your network
* @param ex Expected output values
*/
void setExpected(std::vector<double> ex);
/**
* @brief It updates through back-propagation the weights of the synapsis and
* computes again the output value for <i>epochs</i> times, calling back
@ -171,7 +184,7 @@ namespace neuralpp {
* @brief It sets the input for the network
* @param v Vector of doubles, containing the values to give to your network
*/
void setInput (vector<double>& v);
void setInput (std::vector<double> v);
/**
* @brief It links the layers of the network (input, hidden, output). Don't use unless
@ -194,13 +207,13 @@ namespace neuralpp {
* @param src Source type from which the XML will be loaded (from a file [default] or from a string)
* @throw InvalidXMLException
*/
void train (string xml, source src) throw(InvalidXMLException);
void train (std::string xml, source src) throw(InvalidXMLException);
/**
* @brief Initialize the training XML for the neural network
* @param xml String that will contain the XML
*/
static void initXML (string& xml);
static void initXML (std::string& xml);
/**
* @brief Splits a string into a vector of doubles, given a delimitator
@ -208,7 +221,7 @@ namespace neuralpp {
* @param str String to be splitted
* @return Vector of doubles containing splitted values
*/
static vector<double> split (char delim, string str);
static std::vector<double> split (char delim, std::string str);
/**
* @brief Get a training set from a string and copies it to an XML
@ -222,13 +235,13 @@ namespace neuralpp {
* @param set String containing input values and expected outputs
* @return XML string
*/
static string XMLFromSet (int id, string set);
static std::string XMLFromSet (int id, std::string set);
/**
* @brief Closes an open XML document generated by "initXML" and "XMLFromSet"
* @param xml XML string to be closed
*/
static void closeXML(string& xml);
static void closeXML(std::string& xml);
};
/**
@ -337,8 +350,8 @@ namespace neuralpp {
double actv_val;
double prop_val;
vector< Synapsis > in;
vector< Synapsis > out;
std::vector< Synapsis > in;
std::vector< Synapsis > out;
double (*actv_f)(double);
@ -355,7 +368,7 @@ namespace neuralpp {
* @param out Output synapses
* @param a Activation function
*/
Neuron (vector<Synapsis> in, vector<Synapsis> out,
Neuron (std::vector<Synapsis> in, std::vector<Synapsis> out,
double (*a)(double));
/**
@ -376,13 +389,13 @@ namespace neuralpp {
* @brief It pushes a new input synapsis
* @param s Synapsis to be pushed
*/
void push_in (Synapsis& s);
void push_in (Synapsis s);
/**
* @brief It pushes a new output synapsis
* @param s Synapsis to be pushed
*/
void push_out (Synapsis& s);
void push_out (Synapsis s);
/**
* @brief Change the activation value of the neuron
@ -409,9 +422,9 @@ namespace neuralpp {
double getProp();
/**
* @brief It propagates its activation value to the connected neurons
* @brief Compute the propagation value of the neuron and set it
*/
double propagate();
void propagate();
/**
* @brief Get the number of input synapsis for the neuron
@ -437,7 +450,7 @@ namespace neuralpp {
* you're doing, use NeuralNet instead
*/
class Layer {
vector<Neuron> elements;
std::vector<Neuron> elements;
void (*update_weights)();
double (*actv_f)(double);
@ -456,7 +469,7 @@ namespace neuralpp {
* @param neurons Vector of neurons to be included in the layer
* @param a Activation function
*/
Layer (vector<Neuron>& neurons, double(*a)(double));
Layer (std::vector<Neuron>& neurons, double(*a)(double));
/**
* @brief Redefinition for operator []. It gets the neuron at <i>i</i>
@ -472,16 +485,10 @@ namespace neuralpp {
void link (Layer& l);
/**
* @brief It sets a vector of propagation values to all its neurons
* @param v Vector of values to write as propagation values
* @brief Set the input values for the neurons of the layer (just use it for the input layer)
* @param v Vector containing the input values
*/
void setProp (vector<double>& v);
/**
* @brief It sets a vector of activation values to all its neurons
* @param v Vector of values to write as activation values
*/
void setActv (vector<double>& v);
void setInput (std::vector<double> v);
/**
* @brief It propagates its activation values to the output layers

View file

@ -14,6 +14,8 @@
#include <cstdlib>
#include "neural++.hpp"
using std::vector;
namespace neuralpp {
Layer::Layer(size_t sz, double (*a) (double)) {
for (size_t i = 0; i < sz; i++) {
@ -56,23 +58,16 @@ namespace neuralpp {
}
}
void Layer::setProp(vector < double >&v) {
for (size_t i = 0; i < size(); i++)
void Layer::setInput (vector<double> v) {
for (size_t i = 0; i < size(); i++) {
elements[i].setProp(v[i]);
}
void Layer::setActv(vector < double >&v) {
for (size_t i = 0; i < size(); i++)
elements[i].setActv(v[i]);
}
}
void Layer::propagate() {
for (size_t i = 0; i < size(); i++) {
Neuron *n = &(elements[i]);
n->setProp(n->propagate());
n->setActv(actv_f(n->getProp()));
}
for (size_t i = 0; i < size(); i++)
elements[i].propagate();
}
}

View file

@ -13,6 +13,7 @@
#include <fstream>
#include <sstream>
using namespace std;
#include "neural++.hpp"
#include "Markup.h"
@ -78,9 +79,8 @@ namespace neuralpp {
output->propagate();
}
void NeuralNet::setInput(vector <double>& v) {
input->setProp(v);
input->setActv(v);
void NeuralNet::setInput(vector <double> v) {
input->setInput(v);
}
void NeuralNet::link() {
@ -89,11 +89,12 @@ namespace neuralpp {
}
void NeuralNet::setExpected(double e) {
ex = e;
ex.clear();
ex.push_back(e);
}
double NeuralNet::expected() const {
return ex;
return ex[0];
}
void NeuralNet::updateWeights() {
@ -101,10 +102,6 @@ namespace neuralpp {
for (size_t i = 0; i < output->size(); i++) {
Neuron *n = &(*output)[i];
double prop = 0.0;
for (size_t j = 0; j < n->nIn(); j++)
prop += (n->synIn(j).getWeight() * n->synIn(j).getIn()->getActv());
for (size_t j = 0; j < n->nIn(); j++) {
Synapsis *s = &(n->synIn(j));
@ -112,13 +109,13 @@ namespace neuralpp {
if (ref_epochs - epochs > 0)
out_delta =
(-l_rate) * (getOutput() - expected()) *
df(actv_f, prop) * s->getIn()->getActv() +
df(actv_f, n->getProp()) * s->getIn()->getActv() +
s->momentum(ref_epochs, ref_epochs - epochs) *
s->getPrevDelta();
else
out_delta =
(-l_rate) * (getOutput() - expected()) *
df(actv_f, prop) * s->getIn()->getActv();
df(actv_f, n->getProp()) * s->getIn()->getActv();
s->setDelta(out_delta);
}
@ -148,9 +145,9 @@ namespace neuralpp {
}
}
void NeuralNet::commitChanges(Layer * l) {
for (size_t i = 0; i < l->size(); i++) {
Neuron *n = &(*l)[i];
void NeuralNet::commitChanges(Layer& l) {
for (size_t i = 0; i < l.size(); i++) {
Neuron *n = &(l[i]);
for (size_t j = 0; j < n->nIn(); j++) {
Synapsis *s = &(n->synIn(j));
@ -164,8 +161,8 @@ namespace neuralpp {
void NeuralNet::update() {
while ((epochs--) > 0) {
updateWeights();
commitChanges(output);
commitChanges(hidden);
commitChanges(*output);
commitChanges(*hidden);
propagate();
}
}
@ -183,7 +180,7 @@ namespace neuralpp {
record.epochs = ref_epochs;
record.l_rate = l_rate;
record.ex = ex;
record.ex = ex[0];
if (out.write((char*) &record, sizeof(struct netrecord)) <= 0)
throw NetworkFileWriteException();

View file

@ -13,6 +13,8 @@
#include "neural++.hpp"
using std::vector;
namespace neuralpp {
Neuron::Neuron(double (*a) (double)) {
actv_f = a;
@ -35,11 +37,11 @@ namespace neuralpp {
return out[i];
}
void Neuron::push_in(Synapsis & s) {
void Neuron::push_in(Synapsis s) {
in.push_back(s);
}
void Neuron::push_out(Synapsis & s) {
void Neuron::push_out(Synapsis s) {
out.push_back(s);
}
@ -67,13 +69,15 @@ namespace neuralpp {
return actv_val;
}
double Neuron::propagate() {
double aux = 0;
void Neuron::propagate() {
double aux = 0.0;
for (size_t i = 0; i < nIn(); i++)
aux +=
(in[i].getWeight() * in[i].getIn()->actv_val);
return aux;
setProp(aux);
setActv( actv_f(getProp()) );
}
void Neuron::synClear() {