Line data    Source code

       1             : /*************************************************************************
       2             :  *
       3             :  *  Project
       4             :  *                         _____ _____  __  __ _____
       5             :  *                        / ____|  __ \|  \/  |  __ \
       6             :  *  ___  _ __   ___ _ __ | |  __| |__) | \  / | |__) |
       7             :  * / _ \| '_ \ / _ \ '_ \| | |_ |  ___/| |\/| |  ___/
       8             :  *| (_) | |_) |  __/ | | | |__| | |    | |  | | |
       9             :  * \___/| .__/ \___|_| |_|\_____|_|    |_|  |_|_|
      10             :  *      | |
      11             :  *      |_|
      12             :  *
      13             :  * Copyright (C) Akiel Aries, <akiel@akiel.org>, et al.
      14             :  *
      15             :  * This software is licensed as described in the file LICENSE, which
      16             :  * you should have received as part of this distribution. The terms
      17             :  * among other details are referenced in the official documentation
      18             :  * seen here : https://akielaries.github.io/openGPMP/ along with
      19             :  * important files seen in this project.
      20             :  *
      21             :  * You may opt to use, copy, modify, merge, publish, distribute
      22             :  * and/or sell copies of the Software, and permit persons to whom
      23             :  * the Software is furnished to do so, under the terms of the
      24             :  * LICENSE file. As this is an Open Source effort, all implementations
      25             :  * must be of the same methodology.
      26             :  *
      27             :  *
      28             :  *
      29             :  * This software is distributed on an AS IS basis, WITHOUT
      30             :  * WARRANTY OF ANY KIND, either express or implied.
      31             :  *
      32             :  ************************************************************************/
      33             : 
      34             : /*
      35             :  * Implementation of a Multi-Layered Perceptron Neural Network
      36             :  */
      37             : #include <math.h>
      38             : #include <openGPMP/ml/mlp_net.hpp>
      39             : #include <stdio.h>
      40             : #include <string.h>
      41             : #include <time.h>
      42             : 
      43             : using namespace gpmp::ml;
      44             : 
      45             : /*
      46             :  * Initialize randomly generated values for network's method
      47             :  */
      48           0 : void gpmp::ml::PrimaryMLP::rand_init() {
      49           0 :     srand(4711);
      50             :     // srand((uint64_t)time(NULL));
      51           0 : }
      52             : 
      53             : /* verify the random is an integer */
      54           0 : int64_t gpmp::ml::PrimaryMLP::rand_int(int64_t hi, int64_t low) {
      55           0 :     return rand() % (hi - low + 1) + low;
      56             : }
      57             : 
      58             : /* verify generated random is a real number */
      59           0 : long double gpmp::ml::PrimaryMLP::rand_real(long double low, long double hi) {
      60           0 :     return ((long double)rand() / RAND_MAX) * (hi - low) + low;
      61             : }
      62             : 
      63             : /* PRIMARY MLP CONSTRUCTOR */
      64           0 : gpmp::ml::PrimaryMLP::PrimaryMLP(int64_t nl, int64_t npl[])
      65           0 :     : num_layers(0), layer_ptr(0), _MSE(0.0), _MAE(0.0), _Eta(0.25),
      66           0 :       _Alpha(0.9), _Gain(1.0), _AvgTestError(0.0) {
      67             :     int64_t _LAYER, _NEURON;
      68             : 
      69             :     // create network layers
      70           0 :     num_layers = nl;
      71           0 :     layer_ptr = new layer[nl];
      72             : 
      73             :     // intialize the data of the created network layers
      74           0 :     for (_LAYER = 0; _LAYER < nl; _LAYER++) {
      75             :         // intialize values to neuron struct information
      76           0 :         layer_ptr[_LAYER].num_neurons = npl[_LAYER];
      77           0 :         layer_ptr[_LAYER].neuron_ptr = new neuron[npl[_LAYER]];
      78             : 
      79             :         // intialize date of the neurons of the created network layers
      80           0 :         for (_NEURON = 0; _NEURON < npl[_LAYER]; _NEURON++) {
      81             :             // initialize exit value
      82           0 :             layer_ptr[_LAYER].neuron_ptr[_NEURON].sortir = 1.0;
      83             :             // save the error
      84           0 :             layer_ptr[_LAYER].neuron_ptr[_NEURON].err = 0.0;
      85             : 
      86             :             // check if there is at least 1 layer
      87           0 :             if (_LAYER > 0) {
      88             :                 /* initialize weight, last weight, and saved weight
      89             :                  * values to _LAYER - 1
      90             :                  */
      91           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt =
      92           0 :                     new long double[npl[_LAYER - 1]];
      93             : 
      94           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_last =
      95           0 :                     new long double[npl[_LAYER - 1]];
      96             : 
      97           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_saved =
      98           0 :                     new long double[npl[_LAYER - 1]];
      99             :             }
     100             :             // otherwise
     101             :             else {
     102             :                 /*
     103             :                  * initialize weight, last weight, and saved weight
     104             :                  * to NULL
     105             :                  */
     106           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt = NULL;
     107           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_last = NULL;
     108           0 :                 layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_saved = NULL;
     109             :             }
     110             :         }
     111             :     }
     112           0 : }
     113             : 
     114             : /* PRIMARY MLP DECONSTRUCTOR */
     115           0 : gpmp::ml::PrimaryMLP::~PrimaryMLP() {
     116             :     int64_t _LAYER, _NEURON;
     117             : 
     118             :     // TODO : thread the loops dealing with rudimentary computations
     119             : 
     120           0 :     for (_LAYER = 0; _LAYER < num_layers; _LAYER++) {
     121           0 :         if (layer_ptr[_LAYER].neuron_ptr) {
     122           0 :             for (_NEURON = 0; _NEURON < layer_ptr[_LAYER].num_neurons;
     123             :                  _NEURON++) {
     124           0 :                 if (layer_ptr[_LAYER].neuron_ptr[_NEURON].wt) {
     125           0 :                     delete[] layer_ptr[_LAYER].neuron_ptr[_NEURON].wt;
     126             :                 }
     127             : 
     128           0 :                 if (layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_last) {
     129           0 :                     delete[] layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_last;
     130             :                 }
     131             : 
     132           0 :                 if (layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_saved) {
     133           0 :                     delete[] layer_ptr[_LAYER].neuron_ptr[_NEURON].wt_saved;
     134             :                 }
     135             :             }
     136             :         }
     137           0 :         delete[] layer_ptr[_LAYER].neuron_ptr;
     138             :     }
     139           0 :     delete[] layer_ptr;
     140           0 : }