From e369eac255d44a1b3004fdd651a618ef2d6597fe Mon Sep 17 00:00:00 2001
From: Martin Beseda <martin.beseda@vsb.cz>
Date: Wed, 20 Mar 2019 17:06:14 +0100
Subject: [PATCH] WIP: Fixed memory leak
---
src/Network/NeuralNetwork.cpp | 87 ++++++++++++++++++-----------------
src/Network/NeuralNetwork.h | 2 +-
2 files changed, 45 insertions(+), 44 deletions(-)
diff --git a/src/Network/NeuralNetwork.cpp b/src/Network/NeuralNetwork.cpp
index 1979eaef..b2e00d76 100644
--- a/src/Network/NeuralNetwork.cpp
+++ b/src/Network/NeuralNetwork.cpp
@@ -21,8 +21,8 @@ namespace lib4neuro {
this->neuron_biases = new ::std::vector<double>(0);
this->neuron_bias_indices = new ::std::vector<int>(0);
- this->connection_weights = new ::std::vector<double>(0);
- this->connection_list = ::std::vector<std::shared_ptr<ConnectionFunctionGeneral>>(0);
+// this->connection_weights = new ::std::vector<double>(0);
+// this->connection_list = ::std::vector<std::shared_ptr<ConnectionFunctionGeneral>>(0);
this->inward_adjacency = new ::std::vector<std::vector<std::pair<size_t, size_t>> *>(0);
this->outward_adjacency = new ::std::vector<std::vector<std::pair<size_t, size_t>> *>(0);
@@ -85,10 +85,10 @@ namespace lib4neuro {
this->input_neuron_indices = nullptr;
}
- if (this->connection_weights && this->delete_weights) {
- delete this->connection_weights;
- this->connection_weights = nullptr;
- }
+// if (this->connection_weights && this->delete_weights) {
+// delete this->connection_weights;
+// this->connection_weights = nullptr;
+// }
if (this->neuron_biases && this->delete_biases) {
delete this->neuron_biases;
@@ -462,9 +462,9 @@ namespace lib4neuro {
size_t ci = c.second;
this->neuron_potentials->at(ti) +=
- this->connection_list.at(ci)->eval(*this->connection_weights) * potential;
+ this->connection_list.at(ci)->eval(this->connection_weights) * potential;
- std::cout << " adding input to neuron " << ti << " += " << this->connection_list.at(ci)->eval(*this->connection_weights) << "*" << potential << std::endl;
+ std::cout << " adding input to neuron " << ti << " += " << this->connection_list.at(ci)->eval(this->connection_weights) << "*" << potential << std::endl;
}
}
}
@@ -494,10 +494,10 @@ namespace lib4neuro {
// con_weight_u1u2 = new ConnectionFunctionIdentity();
} else {
if (sct == SIMPLE_CONNECTION_TYPE::NEXT_WEIGHT) {
- weight_idx = this->connection_weights->size();
- this->connection_weights->resize(weight_idx + 1);
+ weight_idx = this->connection_weights.size();
+ this->connection_weights.resize(weight_idx + 1);
} else if (sct == SIMPLE_CONNECTION_TYPE::EXISTING_WEIGHT) {
- if (weight_idx >= this->connection_weights->size()) {
+ if (weight_idx >= this->connection_weights.size()) {
::std::cerr << "The supplied connection weight index is too large!\n" << ::std::endl;
}
}
@@ -544,20 +544,21 @@ namespace lib4neuro {
void NeuralNetwork::copy_parameter_space(std::vector<double> *parameters) {
if (parameters != nullptr) {
- for (unsigned int i = 0; i < this->connection_weights->size(); ++i) {
- (*this->connection_weights).at(i) = (*parameters).at(i);
+ for (unsigned int i = 0; i < this->connection_weights.size(); ++i) {
+ this->connection_weights.at(i) = (*parameters).at(i);
}
for (unsigned int i = 0; i < this->neuron_biases->size(); ++i) {
- (*this->neuron_biases).at(i) = (*parameters).at(i + this->connection_weights->size());
+ (*this->neuron_biases).at(i) = (*parameters).at(i + this->connection_weights.size());
}
}
}
void NeuralNetwork::set_parameter_space_pointers(NeuralNetwork &parent_network) {
- if (this->connection_weights) {
- delete connection_weights;
+ if (!this->connection_weights.empty()) {
+// delete connection_weights;
+ this->connection_weights.clear();
}
if (this->neuron_biases) {
@@ -620,7 +621,7 @@ namespace lib4neuro {
size_t ci = c.second;
this->neuron_potentials->at(ti) +=
- this->connection_list.at(ci)->eval(*this->connection_weights) * potential;
+ this->connection_list.at(ci)->eval(this->connection_weights) * potential;
}
}
}
@@ -689,7 +690,7 @@ namespace lib4neuro {
size_t ci = c.second;
neuron_potential_t = this->neurons->at(ti)->get_last_activation_value( );
- connection_weight = this->connection_list.at(ci)->eval(*this->connection_weights);
+ connection_weight = this->connection_list.at(ci)->eval(this->connection_weights);
this->connection_list.at(ci)->eval_partial_derivative(*this->get_parameter_ptr_weights(),
gradient,
@@ -766,7 +767,7 @@ namespace lib4neuro {
size_t ci = c.second;
neuron_activation_t = this->neurons->at(ti)->get_last_activation_value( );
- connection_weight = this->connection_list.at(ci)->eval(*this->connection_weights);
+ connection_weight = this->connection_list.at(ci)->eval(this->connection_weights);
std::cout << " [backpropagation] value ("<<ti<< "): " << neuron_activation_t << ", scaling: " << scaling_backprog[neuron_idx] << std::endl;
@@ -792,12 +793,12 @@ namespace lib4neuro {
boost::random::mt19937 gen(std::time(0));
// Init weight guess ("optimal" for logistic activation functions)
- double r = 4 * sqrt(6. / (this->connection_weights->size()));
+ double r = 4 * sqrt(6. / (this->connection_weights.size()));
boost::random::uniform_real_distribution<> dist(-r, r);
- for (size_t i = 0; i < this->connection_weights->size(); i++) {
- this->connection_weights->at(i) = dist(gen);
+ for (size_t i = 0; i < this->connection_weights.size(); i++) {
+ this->connection_weights.at(i) = dist(gen);
}
}
@@ -825,7 +826,7 @@ namespace lib4neuro {
void NeuralNetwork::scale_weights(double alpha) {
for(size_t i = 0; i < this->get_n_weights(); ++i){
- this->connection_weights->at( i ) *= alpha;
+ this->connection_weights.at( i ) *= alpha;
}
}
@@ -843,7 +844,7 @@ namespace lib4neuro {
}
size_t NeuralNetwork::get_n_weights() {
- return this->connection_weights->size();
+ return this->connection_weights.size();
}
size_t NeuralNetwork::get_n_biases() {
@@ -878,11 +879,11 @@ namespace lib4neuro {
void NeuralNetwork::write_weights() {
std::cout << "Connection weights: ";
- if (this->connection_weights) {
- for (size_t i = 0; i < this->connection_weights->size() - 1; ++i) {
- std::cout << this->connection_weights->at(i) << ", ";
+ if (!this->connection_weights.empty()) {
+ for (size_t i = 0; i < this->connection_weights.size() - 1; ++i) {
+ std::cout << this->connection_weights.at(i) << ", ";
}
- std::cout << this->connection_weights->at(this->connection_weights->size() - 1) << std::endl;
+ std::cout << this->connection_weights.at(this->connection_weights.size() - 1) << std::endl;
}
}
@@ -895,22 +896,22 @@ namespace lib4neuro {
ofs << "Connection weights: ";
- if (this->connection_weights) {
- for (size_t i = 0; i < this->connection_weights->size() - 1; ++i) {
- ofs << this->connection_weights->at(i) << ", ";
+ if (!this->connection_weights.empty()) {
+ for (size_t i = 0; i < this->connection_weights.size() - 1; ++i) {
+ ofs << this->connection_weights.at(i) << ", ";
}
- ofs << this->connection_weights->at(this->connection_weights->size() - 1) << std::endl;
+ ofs << this->connection_weights.at(this->connection_weights.size() - 1) << std::endl;
}
}
void NeuralNetwork::write_weights(std::ofstream* file_path) {
*file_path << "Connection weights: ";
- if (this->connection_weights) {
- for (size_t i = 0; i < this->connection_weights->size() - 1; ++i) {
- *file_path << this->connection_weights->at(i) << ", ";
+ if (!this->connection_weights.empty()) {
+ for (size_t i = 0; i < this->connection_weights.size() - 1; ++i) {
+ *file_path << this->connection_weights.at(i) << ", ";
}
- *file_path << this->connection_weights->at(this->connection_weights->size() - 1) << std::endl;
+ *file_path << this->connection_weights.at(this->connection_weights.size() - 1) << std::endl;
}
}
@@ -957,7 +958,7 @@ namespace lib4neuro {
::std::cout << std::flush
<< "Number of neurons: " << this->neurons->size() << ::std::endl
<< "Number of connections: " << this->connection_list.size() << ::std::endl
- << "Number of active weights: " << this->connection_weights->size() << ::std::endl
+ << "Number of active weights: " << this->connection_weights.size() << ::std::endl
<< "Number of active biases: " << this->neuron_biases->size() << ::std::endl;
if(this->normalization_strategy) {
@@ -979,7 +980,7 @@ namespace lib4neuro {
ofs << "Number of neurons: " << this->neurons->size() << ::std::endl
<< "Number of connections: " << this->connection_list.size() << ::std::endl
- << "Number of active weights: " << this->connection_weights->size() << ::std::endl
+ << "Number of active weights: " << this->connection_weights.size() << ::std::endl
<< "Number of active biases: " << this->neuron_biases->size() << ::std::endl;
if(this->normalization_strategy) {
@@ -996,7 +997,7 @@ namespace lib4neuro {
void NeuralNetwork::write_stats(std::ofstream* file_path) {
*file_path << "Number of neurons: " << this->neurons->size() << ::std::endl
<< "Number of connections: " << this->connection_list.size() << ::std::endl
- << "Number of active weights: " << this->connection_weights->size() << ::std::endl
+ << "Number of active weights: " << this->connection_weights.size() << ::std::endl
<< "Number of active biases: " << this->neuron_biases->size() << ::std::endl;
if(this->normalization_strategy) {
@@ -1008,12 +1009,12 @@ namespace lib4neuro {
}
}
- std::vector<double> *NeuralNetwork::get_parameter_ptr_biases() {
+ std::vector<double>* NeuralNetwork::get_parameter_ptr_biases() {
return this->neuron_biases;
}
- std::vector<double> *NeuralNetwork::get_parameter_ptr_weights() {
- return this->connection_weights;
+ std::vector<double>* NeuralNetwork::get_parameter_ptr_weights() {
+ return &this->connection_weights;
}
// size_t NeuralNetwork::add_new_connection_to_list(ConnectionFunctionGeneral *con) {
@@ -1236,7 +1237,7 @@ namespace lib4neuro {
this->neuron_potentials = new ::std::vector<double>(0);
this->neuron_bias_indices = new ::std::vector<int>(0);
- this->connection_weights = new ::std::vector<double>(0);
+// this->connection_weights = new ::std::vector<double>(0);
this->connection_list = ::std::vector<std::shared_ptr<ConnectionFunctionGeneral>>(0);
this->inward_adjacency = new ::std::vector<std::vector<std::pair<size_t, size_t>> *>(0);
this->outward_adjacency = new ::std::vector<std::vector<std::pair<size_t, size_t>> *>(0);
diff --git a/src/Network/NeuralNetwork.h b/src/Network/NeuralNetwork.h
index f7cc7a16..9f69eb33 100644
--- a/src/Network/NeuralNetwork.h
+++ b/src/Network/NeuralNetwork.h
@@ -70,7 +70,7 @@ namespace lib4neuro {
/**
*
*/
- std::vector<double> *connection_weights = nullptr;
+ std::vector<double> connection_weights; //= nullptr;
/**
*
--
GitLab