int main(int argc, char **argv) {
int num_comms = 2;
int triple_type = D3CS_NOBACK;
int cut_type = BIPARTITE;
std::string name = "tripartite_flow";
std::cout << "DIRLAP ---------" << std::endl;
{
Network net = Network("data", name, triple_type);
CommDetection comm(net, num_comms, DIRLAP, cut_type, name);
comm.Run();
}
std::cout << "D3C_ONLY ---------" << std::endl;
{
Network net = Network("data", name, triple_type);
CommDetection comm(net, num_comms, D3C_ONLY_DIRLAP, cut_type, name);
comm.Run();
}
std::cout << "TENSOR ---------" << std::endl;
{
Network net = Network("data", name, triple_type);
CommDetection comm(net, num_comms, TENSOR, cut_type, name);
comm.Run();
}
std::cout << "END ---------" << std::endl;
}
Network Network::read(istream &s) {
Network fail = Network(0,0);
if(!s.good()) {
return fail;
}
string keyword;
s >> keyword;
if(keyword != "NETWORK") {
{
return fail;
}
}
s >> keyword;
if(keyword != "input_size") {
{
return fail;
}
}
int input_size;
s >> input_size;
s >> keyword;
if(keyword != "layers") {
return fail;
}
int layers;
s >> layers;
char c = s.get();
if(c != '\n') {
return fail;
}
shared_ptr<Layer> first(Layer::read(s, input_size));
if (first->size() == 0) {
return fail;
}
layers--;
if (layers == 0) {
return Network(input_size, shared_ptr<Layer>(NULL), first);
}
shared_ptr<Layer> current = shared_ptr<Layer>(Layer::read(s, first));
if (current->size() == 0) {
return fail;
}
first->setNextLayer(current);
layers--;
while(layers > 0) {
current->setNextLayer(shared_ptr<Layer>(Layer::read(s, current)));
current = current->nextLayer();
if (current->size() == 0) {
return fail;
}
layers--;
}
return Network(input_size, first, current);
}
void main (){
int gdriver=0,gmode,opc,opc1,opc2,opc3,opc4,opc5,opc6,opc7,opc8,opc9;
float Periodos,Tasa;
initgraph(&gdriver,&gmode," ");
mouse.Inicializar(); //Inicializa Mouse
boton(90,70,500,300,8,15,7,5);
barra_titulo(92,72,499,"Ingreso De Datos - Mac Software");
DibujeIcono(97,73,Network(),16,16);
boton(483,75,495,87,8,15,7,1);DibujeIcono(485,77,Cerrar(),9,9);
marco_grabado(100,98,490,290);
Periodos=CuadroDato(115,"Digite Numero de Periodos","");
Tasa=CuadroDato(185,"Digite Tasa de Interes","");
conv_text(125,253,"NOTA: ");
conv_text(125,268,"EL PERIODO CERO (0) ES EL MOMENTO INICIAL");
getch();
cleardevice();
printf("Periodo= %f Tasa=%f",Periodos,Tasa);
getch();
closegraph();exit(0);restorecrtmode();system("cls");
} //FIN
int main(int argc,char *argv[]){
Network nw=Network();
nw.addCompartment(0,"unknown");
nw.addCompartment(1,"ex");
nw.addCompartment(2,"pm");
nw.addCompartment(3,"cy");
nw.addNode(0, reaction , "R", 100 , 100 , 00, 00 ,0,2);
nw.addNode(1, compound , "c1", 100 , 100 , 300, 00 , 00,1);
nw.addNode(2, compound , "c2", 100 , 100 , 300, 300 , 00,1);
nw.addNode(3, compound , "c3", 100 , 100 , 00, 300 , 00,2);
nw.addNode(4, compound , "c4", 100 , 100 , 200, 300 , 00,2);
nw.addNode(5, compound , "c5", 100 , 100 , 400, 300 , 00,3);
nw.addEdge(0,1,substrate);
nw.addEdge(0,2,substrate);
nw.addEdge(0,3,product);
nw.addEdge(0,4,inhibitor);
nw.addEdge(0,5,catalyst);
nw.dump();
Plugins& pgs=register_plugins();
Layouter l(nw,pgs);
l.stepAddPlugins(0,P_force_adj, P_torque_adj, P_distribute_edges, P_adjust_compartments, P_force_compartments);
l.stepAddEndCondition(0,C_relForceDiff,0.0005);
l.stepAddEndCondition(0,C_temp,10);
l.stepAddPlugins(1,P_fix_compartments);
l.stepAddEndCondition(1,C_iterations,1);
l.stepAddPlugins(2,P_force_adj, P_torque_adj,P_force_nadj, P_distribute_edges, P_adjust_compartments_fixed, P_force_compartments);
l.stepAddEndCondition(2,C_relForceDiff,0.0005);
l.stepAddEndCondition(2,C_temp,10);
l.execute();
printf("finished. [press key]\n");
getc(stdin);
return 0;
}
int main(int argc,char *argv[]) {
if(argc != 5) {
printf("Wrong Number of Arguments\n");
exit(0);
}
//double alpha1 = argv[1]-0.0;
double alpha1, alpha2, beta1, beta2, gamma1, gamma2, phi1, phi2;
alpha1 = alpha2 = atof( argv[1] );
beta1 = 0.35;
beta2 = atof( argv[2] );
gamma1 = gamma2 = 1.0/5.0;
phi1 = phi2 = atof( argv[4]);
int intro_time;
intro_time = atoi( argv[3] );
int num_reps = 5;
Network net = Network("gillespie toy", Network::Undirected);
net.populate(10000);
net.connect_all_nodes();
Gillespie_TwoStrain_Network_Sim sim(&net, alpha1, alpha2, gamma1, gamma2, beta1, beta2, phi1, phi2, intro_time);
for(int i =1; i <= num_reps; i++){
cout << "Simulation number: " << i << endl;
sim.reset();
sim.rand_infect(5, 1);
sim.run_simulation(10000.0);
}
return 0;
}
void main()
{
for (int i = 0; i < 4; i++)
{
PlayerAddresses[i].port = 0;
}
NetDevice = RakNet::RakPeerInterface::GetInstance();
NetDevice->Startup(MaxClients, &RakNet::SocketDescriptor(ServerPort, 0), 1);
Print("Starting the server...");
Skip();
NetDevice->SetMaximumIncomingConnections(MaxClients);
while(1)
{
TimeDelta = timeGetTime() - Start;
Start = timeGetTime();
for (int i = 0; i < 1000; i++)
{
std::cout << "";
}
Timers();
Network();
Game();
}
RakNet::RakPeerInterface::DestroyInstance(NetDevice);
}
int main(void) {
int i;
int64_t llTime;
llTime = GetTime();
for (i = 2; i <= 64; i += 2) {
printf("R(%d)=%.15f(%dms)\n", i, Network(i), (int) (GetTime() - llTime));
}
return 0;
}
// runs a seasonally forced gillespie simulation on a exponential network
int main(int argc,char *argv[]) {
if(argc != 9) {
printf("Wrong Number of Arguments\n");
exit(0);
}
double alpha1, alpha2, beta1_max, beta2_max, gamma1, gamma2, phi1, phi2, eta1, eta2;
alpha1 = alpha2 = atof( argv[1] );
beta1_max = atof( argv[2] );
beta2_max = atof( argv[3] );
eta1 = eta2 = 1.0/2.62;
gamma1 = gamma2 = 1.0/3.38;
phi1 = phi2 = atof( argv[4]);
int intro_time, start_ind, shift;
intro_time = atoi( argv[5] );
start_ind = atoi( argv[6] );
shift = atoi( argv[7] );
string network_type = argv[8];
int num_reps = 4000;
Network net = Network("gillespie toy", Network::Undirected);
net.populate(10000);
for(int i =1; i <= num_reps; i++){
net.clear_edges();
if(network_type == "exp"){
net.rand_connect_exponential(0.06453487);
} else if(network_type == "unif"){
vector<int> degrees(10000, 16);
net.rand_connect_explicit(degrees);
} else{
cerr << "Unrecognized network type" << endl;
}
// cout << net.mean_deg() << endl;
Gillespie_SEIR_TwoStrain_Network sim(&net, alpha1, alpha2, eta1, eta2, gamma1, gamma2, beta1_max, beta2_max, phi1, phi2, intro_time, start_ind, shift);
cout << "Simulation number: " << i << endl;
sim.reset();
sim.rand_infect(5, 1);
sim.run_simulation(10000.0);
}
return 0;
}
int main(int argc,char *argv[]) {
if(argc != 10) {
printf("Wrong Number of Arguments\n");
exit(0);
}
//double alpha1 = argv[1]-0.0;
double beta_vax, beta_dis, gamma_vax, gamma_dis;
int init_inf, init_vax;
beta_vax = atof( argv[1] ); // vaccine transmissibility
beta_dis = atof( argv[2] ); // disease transmissibility
gamma_vax = atof( argv[3] ); // vaccine recovery rate
gamma_dis = atof( argv[4] ); // disease recovery rate
init_vax = atoi( argv[5] ); // number initially vaccinated
init_inf = atoi( argv[6] ); // number initially infected
string network_type = argv[7] ;
string vax_type = argv[8];
int num_reps = atoi( argv[9]);
Network net = Network("gillespie toy", Network::Undirected);
net.populate(10000);
for(int i =1; i <= num_reps; i++){
net.clear_edges();
if(network_type == "er"){
net.erdos_renyi(10);
} else if(network_type == "exp"){
net.rand_connect_exponential(0.1);
} else if(network_type == "unif"){
vector<int> degrees(10000, 10);
net.rand_connect_explicit(degrees);
} else{
cerr << "Unrecognized network type" << endl;
}
Gillespie_trans_vaccine sim(&net, gamma_vax, gamma_dis, beta_vax, beta_dis, vax_type);
cout << "Simulation number: " << i << endl;
sim.reset();
sim.run_simulation(10000.0, init_vax, init_inf);
}
return 0;
}
int main(int argc,char *argv[]){
Network nw=Network();
nw.addCompartment(0,"unknown");
nw.addNode(0, reaction , "R1", 100 , 100 , 00, 00 , 00,00);
nw.addNode(1, compound , "c1", 100 , 100 , 300, 00 , 00,00);
nw.addNode(2, compound , "c2", 100 , 100 , 300, 300 , 00,00);
nw.addNode(3, compound , "c3", 100 , 100 , 00, 300 , 00,00);
nw.addNode(4, compound , "c4", 100 , 100 , 200, 300 , 00,00);
nw.addNode(5, compound , "c5", 100 , 100 , 400, 300 , 00,00);
nw.addNode(6, reaction , "R2", 100 , 100 , 300, -300 , 00,00);
nw.addNode(7, compound , "c6", 100 , 100 , 500, -300 , 00,00);
nw.addNode(8, compound , "c7", 100 , 100 , 500, 00 , 00,00);
nw.addNode(9, compound , "c8", 100 , 100 , 500, 200 , 00,00);
nw.addNode(10, compound , "c9", 100 , 100 , 300, 200 , 00,00);
nw.addEdge(0,1,substrate);
nw.addEdge(0,2,substrate);
nw.addEdge(0,3,product);
nw.addEdge(0,4,inhibitor);
nw.addEdge(0,5,catalyst);
nw.addEdge(6,1,product);
nw.addEdge(6,7,substrate);
nw.addEdge(8,9,directed);
nw.addEdge(9,10,directed);
nw.dump();
Plugins& pgs=register_plugins();
Layouter l(nw,pgs);
l.stepAddPlugins(0,P_force_adj, P_torque_adj, P_distribute_edges, P_force_nadj, P_limit_mov);
l.stepAddEndCondition(0,C_relForceDiff,0.0005);
l.execute();
printf("finished. [press key]\n");
getc(stdin);
return 0;
}
void BPNN_Recommendation::recommendArticleByBP()
{
cout << "Compute NN model. " << endl;
//first compute the model
int neurons[3] = {20, 20, NodeNum};
BPLearning NN(
Network(NodeNum,3,new SigmodFunction(2),neurons),
1,0.0
);
ifstream in("NetModel//netWorkOutput.txt"); //netWork outPut
if(in.fail())
{
cout << "ERROR: fail to open file netWorkOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.output.size(); i++)
in >> NN.network.output[i];
in.close();
in.open("NetModel//laySum.txt"); // lay sum
if(in.fail())
{
cout << "ERROR: fail to open file laySum.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].sum.size(); k++)
{
in >> NN.network.layer[i].sum[k];
}
in.close();
in.open("NetModel//layOutput.txt"); // lay output
if(in.fail())
{
cout << "ERROR: fail to open file layOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].output.size(); k++)
in >> NN.network.layer[i].output[k];
in.close();
in.open("NetModel//layNeuronOutput.txt"); // lay neuron output
if(in.fail())
{
cout << "ERROR: fail to open file layNeuronOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
in >> NN.network.layer[i].neuron[k].output;
in.close();
in.open("NetModel//layNeuronThreshold.txt"); // lay neuron threshold
if(in.fail())
{
cout << "ERROR: fail to open file layNeuronThreshold.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
in >> NN.network.layer[i].neuron[k].threshold;
in.close();
in.open("NetModel//layNeuronWeight.txt"); // lay neuron weight
if(in.fail())
{
cout << "ERROR: fail to open file layNeuronWeight.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
for(int j = 0; j < NN.network.layer[i].neuron[k].weight.size(); j++)
in >> NN.network.layer[i].neuron[k].weight[j];
in.close();
cout << "NN model completed! " << endl;
//get input
ReadXYintoFile("xy.txt");
//figure weight
for(int i = 0; i < m_userList.size(); ++i){
//input
for(int k = 0; k < m_userList[i]->alternativeList.size(); k++)
{
m_userList[i]->alternativeList[k].weight[FINAL] = m_x * m_userList[i]->alternativeList[k].weight[SOCIAL] / m_userList[i]->maxWeght[SOCIAL] +
m_y * m_userList[i]->alternativeList[k].weight[PERSONAL] / m_userList[i]->maxWeght[PERSONAL] +
(1 - m_x - m_y) * m_userList[i]->alternativeList[k].weight[ITEM] / m_userList[i]->maxWeght[ITEM];
//max
if(m_userList[i]->maxWeght[FINAL] < m_userList[i]->alternativeList[k].weight[FINAL])
m_userList[i]->maxWeght[FINAL] = m_userList[i]->alternativeList[k].weight[FINAL];
m_userList[i]->maxWeght[FINAL] = (m_userList[i]->maxWeght[FINAL] == 0? 1.0 : m_userList[i]->maxWeght[FINAL]);
}
}
vector <double> curI;
double accu = 0.0;
//make input and figure output
for(int i = 1; i <= CASENUM; ++i){
curI.clear();
//input
for(int k = 0; k < m_userList[i - 1]->alternativeList.size(); k++)
{
m_userList[i - 1]->alternativeList[k].weight[FINAL] /= m_userList[i - 1]->maxWeght[FINAL];
curI.push_back(m_userList[i - 1]->alternativeList[k].weight[FINAL]);
}
while(curI.size() < NodeNum)
curI.push_back(0.0);
//figure output
vector <double> outp = NN.calc(curI);
/*
if(i == 1)
{
cout << "size: " << outp.size() << endl;
for(int k = 0; k < outp.size(); k++)
cout << outp[k] << " ";
cout << endl;
//break;
}
cout << endl << endl;
*/
//compute weight
for(int m = 0; m < m_userList[i - 1]->alternativeList.size(); m++)
{
outp[m] = (1.0 - outp[m] > 0 ? 1.0 - outp[m] : outp[m] - 1.0);
//if(i == 1)
//cout << outp[m] << " ";
//cout << endl << endl;
m_userList[i - 1]->alternativeList[m].weight[FINAL] = 10.0 - outp[m];
}
/*
if(i == 1)
{
for(int k = 0; k < m_userList[i - 1]->alternativeList.size(); k++)
cout << m_userList[i - 1]->alternativeList[k].weight[FINAL] << " ";
cout << endl;
break;
}
*/
//figure accuracy
vector<WeightArticle> v = getTopN_Y(m_userList[i - 1]->alternativeList, FINAL, TOP_N);
/*
if(i == 1)
{
cout << "v" << endl;
for(int k = 0; k < v.size(); k++)
cout << v[k].id << " " << v[k].weight[FINAL] << endl;
}
*/
int commom = 0;
for(int k = 0; k < m_answer[i - 1].size(); k++)
{
for(int j = 0; j < m_answer[i - 1].size(); j++)
{
if(v[k].id == m_answer[i - 1][j])
commom++;
}
}
accu += (double)commom / m_answer[i - 1].size();
}
accu /= m_userList.size();
cout << "NN model accuracy: " << accu << endl;
}
//**************************
void Box(int x,int y,int x1,int y1,char Titulo []){
boton(x,y,x1,y1,8,15,7,5);marco_grabado(x+10,y+28,x1-10,y1-10);
barra_titulo(x+2,y+2,x1-1,Titulo);DibujeIcono(x+7,y+3,Network(),16,16);
boton(x1-17,y+5,x1-5,y+17,8,15,7,1);DibujeIcono(x1-15,y+7,Cerrar(),9,9);
}
NetworkTest()
:n_(Network())
{
}
TcpConnection::TcpConnection(char* ip, int port) {
struct sockaddr_in myaddr, remaddr;
int fd, slen = sizeof(remaddr);
char buf[BUFLEN]; /* message buffer */
int recvlen; /* # bytes in acknowledgement message */
srand((int) time(NULL));
bool flags[] = { false, false, false, false, false, false, false, true, false };
seqNo = rand() % 10000;
TcpPacket synPacket(seqNo, rand(), flags, 1u, time(0), PACKET_SIZE, nullptr);
/* Initialize Network */
Network network = Network();
/* create a socket */
if ((fd = socket(AF_INET, SOCK_DGRAM, 0)) == -1)
printf("socket created\n");
/* bind it to all local addresses and pick any port number */
memset((char *) &myaddr, 0, sizeof(myaddr));
myaddr.sin_family = AF_INET;
myaddr.sin_addr.s_addr = htonl(INADDR_ANY);
myaddr.sin_port = htons(0);
if (bind(fd, (struct sockaddr *)&myaddr, sizeof(myaddr)) < 0) {
perror("bind failed");
return;
}
/* now define remaddr, the address to whom we want to send messages */
/* For convenience, the host address is expressed as a numeric IP address */
/* that we will convert to a binary format via inet_aton */
memset((char *) &remaddr, 0, sizeof(remaddr));
remaddr.sin_family = AF_INET;
remaddr.sin_port = htons(SERVICE_PORT);
if (inet_pton(remaddr.sin_family, SERVER, &remaddr.sin_addr) == 0) {
fprintf(stderr, "inet_pton() failed\n");
exit(1);
}
/* Set socket timeout */
struct timeval tv;
tv.tv_sec = 0;
tv.tv_usec = 100000;
if (setsockopt(fd, SOL_SOCKET, SO_RCVTIMEO, (char *) &tv, sizeof(tv)) < 0) {
perror("socket timeout set failed");
exit(1);
}
printf("Sending packet to %s port %d\n", SERVER, SERVICE_PORT);
if (sendto(fd, synPacket.buf, PACKET_SIZE, 0, (struct sockaddr *)&remaddr, slen) == -1) {
perror("sendto");
exit(1);
}
printf("syn packet sent\n");
/* now receive an acknowledgement from the server */
recvlen = recvfrom(fd, buf, BUFLEN, 0, (struct sockaddr *)&remaddr, &slen);
if (recvlen >= 0) {
/* check if ACK & SYN flag set in response*/
bool* Aflags = TcpPacket::getFlags(buf);
if (*(Aflags + ACKBIT) && *(Aflags + SYNBIT)) {
printf("Ack received\n");
char* seqno = TcpPacket::getBytes(buf, 0, SEQUENCE_SIZE);
int AseqNo = atoi(seqno);
char* ackno = TcpPacket::getBytes(buf, SEQUENCE_SIZE, ACK_SIZE);
int ano = atoi(ackno);
if (ano != seqNo + 1) {
perror("seq error mismtach");
exit(1);
}
/* construct ack for the packet just received*/
bool flags[] = { false, false, false, false, true, false, false, false, false };
seqNo++;
TcpPacket ackPacket(seqNo, AseqNo + 1, flags, 1u, time(0), PACKET_SIZE, nullptr);
if (sendto(fd, ackPacket.buf, PACKET_SIZE, 0, (struct sockaddr *)&remaddr, slen) == -1) {
perror("ackPacket");
exit(1);
}
printf("Three way handshake complete\n");
/* Free memory */
free(seqno);
free(ackno);
} else {
perror("Ack not received");
exit(1);
}
free(Aflags);
}
closesocket(fd);
}
Network::Network(const std::string address, const quint16 port):
m_hostAddress(m_hostAddress.fromStdString(address)),
m_hostPort(port)
{
Network();
}
Network::Network() {
Network(false,false,false);
}
void BPNN_Recommendation::training()
{
ReadXYintoFile("xy.txt");
//figure weight
for(int i = 0; i < m_userList.size(); ++i){
//input
for(int k = 0; k < m_userList[i]->alternativeList.size(); k++)
{
m_userList[i]->alternativeList[k].weight[FINAL] = m_x * m_userList[i]->alternativeList[k].weight[SOCIAL] / m_userList[i]->maxWeght[SOCIAL] +
m_y * m_userList[i]->alternativeList[k].weight[PERSONAL] / m_userList[i]->maxWeght[PERSONAL] +
(1 - m_x - m_y) * m_userList[i]->alternativeList[k].weight[ITEM] / m_userList[i]->maxWeght[ITEM];
//max
if(m_userList[i]->maxWeght[FINAL] < m_userList[i]->alternativeList[k].weight[FINAL])
m_userList[i]->maxWeght[FINAL] = m_userList[i]->alternativeList[k].weight[FINAL];
m_userList[i]->maxWeght[FINAL] = (m_userList[i]->maxWeght[FINAL] == 0? 1.0 : m_userList[i]->maxWeght[FINAL]);
}
}
vector <vector<double>> input, output;
vector <double> curI, curO;
//make input and output
for(int i = 1; i <= CASENUM; ++i){
curI.clear();
curO.clear();
//input
for(int k = 0; k < m_userList[i - 1]->alternativeList.size(); k++)
{
m_userList[i - 1]->alternativeList[k].weight[FINAL] /= m_userList[i - 1]->maxWeght[FINAL];
curI.push_back(m_userList[i - 1]->alternativeList[k].weight[FINAL]);
//output
bool isNum1 = false;
for(int m = 0; m < m_answer[i - 1].size(); m++)
{
if(m_userList[i - 1]->alternativeList[k].id == m_answer[i - 1][m])
{
curO.push_back(1.0);
isNum1 = true;
break;
}
}
if(!isNum1)
curO.push_back(0.0);
}
while(curI.size() < NodeNum)
curI.push_back(0.0);
input.push_back(curI);
while(curO.size() < NodeNum)
curO.push_back(0.0);
output.push_back(curO);
}
//printNNInput(input);
//printNNOutput(output);
//int neurons[3] = {8, 8, NodeNum};
int neurons[3] = {20, 20, NodeNum};
BPLearning NN(
Network(NodeNum,3,new SigmodFunction(2),neurons),
1,0.0
);
//NN.run(input[0],output[0]);
for(int i = 0;i < TIME; ++i)
{
double error = NN.runAll(input,output);
printf("case %d's error is %lf\n", i, error);
if(error < ACCURACY)
break;
}
/********************save***********************/
ofstream ou("NetModel//netWorkOutput.txt"); //netWork outPut
if(ou.fail())
{
cout << "ERROR: fail to open file netWorkOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.output.size(); i++)
ou << NN.network.output[i] << endl;
ou.close();
ou.open("NetModel//laySum.txt"); // lay sum
if(ou.fail())
{
cout << "ERROR: fail to open file laySum.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].sum.size(); k++)
{
ou << NN.network.layer[i].sum[k] << endl;
}
ou.close();
ou.open("NetModel//layOutput.txt"); // lay output
if(ou.fail())
{
cout << "ERROR: fail to open file layOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].output.size(); k++)
ou << NN.network.layer[i].output[k] << endl;
ou.close();
ou.open("NetModel//layNeuronOutput.txt"); // lay neuron output
if(ou.fail())
{
cout << "ERROR: fail to open file layNeuronOutput.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
ou << NN.network.layer[i].neuron[k].output << endl;
ou.close();
ou.open("NetModel//layNeuronThreshold.txt"); // lay neuron threshold
if(ou.fail())
{
cout << "ERROR: fail to open file layNeuronThreshold.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
ou << NN.network.layer[i].neuron[k].threshold << endl;
ou.close();
ou.open("NetModel//layNeuronWeight.txt"); // lay neuron weight
if(ou.fail())
{
cout << "ERROR: fail to open file layNeuronWeight.txt..." << endl;
return;
}
for(int i = 0; i < NN.network.layer.size(); i++)
for(int k = 0; k < NN.network.layer[i].neuron.size(); k++)
for(int j = 0; j < NN.network.layer[i].neuron[k].weight.size(); j++)
ou << NN.network.layer[i].neuron[k].weight[j] << endl;
ou.close();
cout << "NN model completed! " << endl;
/*
//test
vector <double> data;
for(int i = 0; i < input[0].size(); i++)
data.push_back(input[0][i]);
vector <double> res = NN.calc(data);
for(int i=0;i<(int)res.size();++i)
cout<< res[i] << " ";
cout << endl;
*/
}
