/*--------------------------------------------------------------------------
Author: Christopher L Buckley
Institute: Centre for Computational Neuroscience and Robotics
University of Sussex
FcondNetworkmer, Brighton BN1 9QJ, UK
email to: c.l.buckley@sussex.ac.uk
version: 2009-11-02
--------------------------------------------------------------------------*/
#ifndef fullMGCRate_H
#define fullMGCRate_H
//w[i][j] source i and target j
class fullMGCRate {
public:
list<neuron *> neurs;
list<synapse *> syns;
list<neuron *>::iterator niter;
list<synapse *>::iterator siter;
NeuroSynRate **theNeurons;
Emptysynapse **theSynapses;
PoissonRateNeuron **theORNs;
Emptysynapse **theORNSynapses;
DCInput **directInput;
NeuronModel *model;
rk65n *machine;
double *x, *xn;
double dt, dtx;
int N;
Array2D<double> mWeights;
Array1D<double> mSeqArray;
Array1D<double> vrestMod;
Array1D<double> mCCBias;
void GenNetwork(void);
fullMGCRate();
~fullMGCRate();
void generateNetwork();
void generateConnect();
void init();
void enable();
void PrintThetaValues();
void PrintWeights();
void PrintMaxEig();
Array1D<double> run(double tme, double inputCurr, Array1D<int> inputStart);
double TransferInv(double in, double *params);
double Transfer(double in, double *params);
double TransferDeriv(double in, double *params);
Array2D<double> getJacobian(void);
void SetWeights(Array2D<double> weights, Array1D<double> SeqArray);
Array2D<double> SetCritical(double epsilon);
Array1D<double> SetCC();
void ScaleWeights(double factor);
void ScaleBias(double factor);
double GetMaxEig();
};
fullMGCRate::fullMGCRate() {
dt = 0.0001;
dtx = 0.0;
N = 0;
directInput = new DCInput*[NoORNs];
theNeurons = new NeuroSynRate*[NoTotal];
theORNs = new PoissonRateNeuron*[NoORNs];
//Set up LNs and PNs
for (int i = 0; i < NoTotal; i++) {
if (i < NoFast)
theNeurons[i] = new NeuroSynRate(i, PARAMS_LNFAST);
if(i>=NoFast && i<NoLNs)
theNeurons[i] = new NeuroSynRate(i, PARAMS_LN);
if(i>=NoLNs)
theNeurons[i] = new NeuroSynRate(i, PARAMS_PN);
neurs.push_back(theNeurons[i]);
// directInput[i] = new DCInput(theNeurons[i], 1.0);
}
//The ORNs
for (int i = 0; i < NoORNs; i++) {
theORNs[i] = new PoissonRateNeuron(i, PARAMS_ORN);
neurs.push_back(theORNs[i]);
directInput[i] = new DCInput(theORNs[i], 1.0);
}
//set up AL connections
theSynapses = new Emptysynapse*[NoTotal * NoTotal];
for (int i = 0; i < NoTotal; i++) {
for (int j = 0; j < NoTotal; j++) {
if (i != j) {
theSynapses[j * NoTotal + i] = new Emptysynapse(theNeurons[i],
theNeurons[j], 0.0);
}
}
}
//set up ORN-AL connections
theORNSynapses = new Emptysynapse*[NoORNs + NoORNs * NoPNs];
for (int i = 0; i < NoORNs; i++) {
theORNSynapses[i] = new Emptysynapse(theORNs[i], theNeurons[i+stimNum],
-ORN2LNWeight);
}
for (int i = 0; i < NoORNs; i++) {
for (int j = NoLNs; j < NoTotal; j++) {
theORNSynapses[NoORNs + (j - NoLNs) * NoORNs + i]
= new Emptysynapse(theORNs[i], theNeurons[j],
-ORN2PNWeight);
}
}
enable();
}
fullMGCRate::~fullMGCRate() {
list<neuron *>::iterator i;
list<synapse *>::iterator j;
for (i = neurs.begin(); i != neurs.end(); i++) {
delete *i;
}
for (j = syns.begin(); j != syns.end(); j++) {
delete *j;
}
delete[] theNeurons;
delete[] theORNs;
delete[] directInput;
}
void fullMGCRate::enable() {
model = new NeuronModel(&neurs, &syns, N, cerr);
x = new double[N];
xn = new double[N];
machine = new rk65n(N, rk65_MINDT, rk65_eps, rk65_absEps, rk65_relEps);
}
void fullMGCRate::ScaleWeights(double factor) {
for (int j = 0; j < NoTotal; j++) {
for (int i = 0; i < NoTotal; i++) {
if (i != j)
theSynapses[j * NoTotal + i]->set_gsyn(mWeights[i][j] * factor);
}
}
}
void fullMGCRate::ScaleBias(double factor) {
for (int i = 0; i < NoTotal; i++)
theNeurons[i]->p[11] = (theNeurons[i]->p[11] * factor);
}
void fullMGCRate::SetWeights(Array2D<double> weights, Array1D<double> SeqArray) {
mWeights = weights;
mSeqArray = SeqArray;
//this code alters the wigtsh absed on the rest potetial.
Array1D<double> setter(NoTotal, 0.0);
vrestMod = setter;
for (int i = 0; i < NoTotal; i++) {
double Iin = TransferInv(theNeurons[i]->p[15] * mSeqArray[i],
theNeurons[i]->p);
vrestMod[i] = a * Iin * Iin + b * Iin + c;
}
for (int i = 0; i < NoTotal; i++) {
theNeurons[i]->SetVrest(vrestMod[i]);
}
//cout << endl << "All the sme rest p[oetials" << endl;
for (int j = 0; j < NoTotal; j++) {
for (int i = 0; i < NoTotal; i++) {
if (i != j)
theSynapses[j * NoTotal + i]->set_gsyn(mWeights[i][j]);
}
}
}
void fullMGCRate::PrintWeights() {
cout << "The weighst are: (rate)" << endl;
for (int j = 0; j < NoTotal; j++) {
for (int i = 0; i < NoTotal; i++) {
cout << mWeights[i][j] * (vrestMod[j] - theNeurons[i]->p[12])
<< " ";
}
cout << endl;
}
cout << endl;
}
Array1D<double> fullMGCRate::SetCC() {
Array1D<double> ccBias(NoTotal, 0.0);
vector<double> summer(NoTotal, 0);
for (int j = 0; j < NoTotal; j++)
for (int i = 0; i < NoTotal; i++)
summer[j] = summer[j] + mWeights[i][j] * (vrestMod[j]
- theNeurons[i]->p[12]) * mSeqArray[i];
for (int i = 0; i < NoTotal; i++) {
ccBias[i] = TransferInv(theNeurons[i]->p[15] * mSeqArray[i],
theNeurons[i]->p) - summer[i];
theNeurons[i]->p[11] = ccBias[i];
theNeurons[i]->p[18] = mSeqArray[i];
}
mCCBias = ccBias;
return ccBias;
}
void fullMGCRate::PrintThetaValues() {
cout << "The theta values are (rate):" << endl;
for (int i = 0; i < NoTotal; i++) {
cout << mCCBias[i] << " ";
}
cout << endl;
}
Array2D<double> fullMGCRate::SetCritical(double epsilon) {
Array2D<double> modWeigths(NoTotal, NoTotal, 0.0);
for (int i = 0; i < NoTotal; i++) {
for (int j = 0; j < NoTotal; j++) {
modWeigths[i][j] = mWeights[i][j] * (vrestMod[j]
- theNeurons[i]->p[12]);
}
}
double maxeig = GetMaxEig();
for (int i = 0; i < NoTotal; i++)
for (int j = 0; j < NoTotal; j++)
mWeights[i][j] = mWeights[i][j] * theNeurons[i]->p[15] / (maxeig)
* (epsilon);
SetWeights(mWeights, mSeqArray);
return mWeights;
}
double fullMGCRate::GetMaxEig() {
Array2D<double> lWeigths(NoTotal, NoTotal, 0.0);
for (int i = 0; i < NoTotal; i++) {
for (int j = 0; j < NoTotal; j++) {
double gammaDeriv = TransferDeriv(TransferInv(theNeurons[j]->p[15]
* mSeqArray[j], theNeurons[j]->p), theNeurons[j]->p);
lWeigths[i][j] = mWeights[i][j] * (vrestMod[j]
- theNeurons[i]->p[12]) * gammaDeriv;
}
}
JAMA::Eigenvalue<double> eigenDecomposition(lWeigths);
TNT::Array1D<double> eigenValues;
eigenDecomposition.getRealEigenvalues(eigenValues);
double maxeig = -100000.0;
for (int i = 0; i < NoTotal; i++) {
if (eigenValues[i] > maxeig)
maxeig = eigenValues[i];
}
return maxeig;
}
void fullMGCRate::PrintMaxEig() {
double lMaxEig = GetMaxEig();
cout << " The max eig is:" << lMaxEig << endl;
}
void fullMGCRate::init() {
dt = 0.0001;
dtx = 0.0;
N = 0;
int counter = 0;
int counter2 = 0;
for (niter = neurs.begin(); niter != neurs.end(); niter++) {
double temp[1];
if (counter < NoTotal) {
temp[0] = mSeqArray[counter2];
counter2++;
} else {
temp[0] = 0;
}
(*niter)->init(x, temp);
counter = counter + 1;
}
for (int i = 0; i < NoORNs; i++)
directInput[i]->set_I(0.0);
}
Array2D<double> fullMGCRate::getJacobian(void) {
Array2D<double> jacobian(NoTotal, NoTotal, 0.0);
for (int i = 0; i < NoTotal; i++) {
for (int j = 0; j < NoTotal; j++) {
double gammaDeriv = TransferDeriv(TransferInv(theNeurons[j]->p[15]
* mSeqArray[j], theNeurons[j]->p), theNeurons[j]->p);
jacobian[i][j] = mWeights[i][j] * (vrestMod[j]
- theNeurons[i]->p[12]) * gammaDeriv;
if (i == j) {
jacobian[i][j] = -theNeurons[j]->p[15];
}
}
}
JAMA::Eigenvalue<double> eigenDecomposition(jacobian);
TNT::Array1D<double> eigenValues;
eigenDecomposition.getRealEigenvalues(eigenValues);
// cout << "Its eignevalue is:";
double marker = -100000.0;
for (int i = 0; i < NoTotal; i++) {
if (eigenValues[i] > marker)
marker = eigenValues[i];
}
// cout << marker << " ";
// cout << endl;
// cout << flush;
return jacobian;
}
Array1D<double> fullMGCRate::run(double tme, double inputCurr,
Array1D<int> inputStart) {
Array1D<double> endPoints(NoTotal * 2, 0.0);
stringstream name;
char thename[80];
ofstream NSRateDataN, NSRateDataS, NSRateDataON, NSRateDataOS;
NSRateDataN.precision(10);
NSRateDataS.precision(10);
NSRateDataON.precision(10);
NSRateDataOS.precision(10);
name.clear();
double *tmp;
x[0] = 0;
double factor;
bool once = true;
if (doneFileCreate) {
if (Manual) {
name << globalName << "RateF.dat" << ends;
name >> thename;
NSRateDataN.open(thename);
name.clear();
/*
name << globalName << "RateS.dat" << ends;
name >> thename;
NSRateDataS.open(thename);
name.clear();
name << globalName << "RateOS.dat" << ends;
name >> thename;
NSRateDataOS.open(thename);
name.clear();
name << globalName << "RateON.dat" << ends;
name >> thename;
NSRateDataON.open(thename);
name.clear();
*/
} else {
name << "RateF.dat" << ends;
name >> thename;
NSRateDataN.open(thename);
name.clear();
/*
name << "RateS.dat" << ends;
name >> thename;
NSRateDataS.open(thename);
name.clear();
name << "RateOS.dat" << ends;
name >> thename;
NSRateDataOS.open(thename);
name.clear();
name << "RateON.dat" << ends;
name >> thename;
NSRateDataON.open(thename);
name.clear();
*/
}
}
while (x[0] < tme) {
if (Gradual) {
if (int(x[0] * 10) % int(scaleBin * 10) == 0 && x[0] < scaleTime) {
factor = (scaleBin + x[0]) / scaleTime;
ScaleWeights(0.5 + 0.4 * factor);
}
if (int((x[0] - scaleStart2) * 10) % int(scaleBin * 10) == 0
&& x[0] > scaleStart2 && x[0] < (scaleStart2 + scaleTime2)) {
factor = (scaleBin + x[0] - scaleStart2) / scaleTime2;
ScaleWeights(0.9 + 0.1 * factor);
}
}
for (int i = 0; i < NoORNs; i++) {
if (x[0] < IMPULSESTART || x[0] > IMPULSESTART + IMPULSEDUR)
{
directInput[i]->set_I(0.0);
// directInput[NoTotal-i-1]->set_I(0.0);
// for (int j = NoLNs; j < NoTotal; j++)
// directInput[j]->set_I(0.0);
}
else
{
directInput[i]->set_I(inputCurr);
/*
if(doPatterned)
{
if (int(x[0] * 10) % 6000 >3000)
directInput[i]->set_I(inputCurr);
else
directInput[i]->set_I(0);
}
else
directInput[i]->set_I(inputCurr);
*/
// for (int j = NoLNs; j < NoTotal; j++)
// directInput[j]->set_I(inputCurr*ORN2PNWeight);
}
}
for (int i = 0; i < NoTotal; i++) {
if (isnan(theNeurons[i]->E(x))) {
cerr << "nan encountered!" << endl;
exit(1);
}
}
double tdt = 0.0;
while (abs(tdt - 0.1) > 1e-9) {
dt = min(dt, 0.1 - tdt);
dtx = machine->integrate(x, xn, model, dt);
dtx = min(dtx, 2.0 * dt);
tmp = x;
x = xn;
xn = tmp;
tdt += dt;
dt = dtx;
}
//out to screen progress
if (int(x[0] * 10) % 5000 == 0)
cout << x[0] << endl;
if (int(x[0] * 10) % 10 == 0)
{
if (doneFileCreate && x[0] > OutTime) {
NSRateDataN << x[0];
// NSRateDataS << x[0];
// NSRateDataON << x[0];
// NSRateDataOS << x[0];
for (int i = 0; i < NoTotal; i++) {
NSRateDataN << " " << theNeurons[i]->F(x);
// NSRateDataS << " " << theNeurons[i]->S(x);
}
for (int i = 0; i < NoORNs; i++) {
// NSRateDataON << " " << theORNs[i]->F(x);
// NSRateDataOS << " " << theORNs[i]->S(x);
}
NSRateDataN << endl;
// NSRateDataS << endl;
// NSRateDataON << endl;
// NSRateDataOS << endl;
}
}
if (once && x[0] > IMPULSESTART - 0.1) {
for (int i = 0; i < NoTotal; i++) {
endPoints[i] = theNeurons[i]->F(x);
}
once = false;
}
}
if (doneFileCreate) {
NSRateDataN.close();
// NSRateDataS.close();
// NSRateDataON.close();
// NSRateDataOS.close();
}
for (int i = NoTotal; i < 2 * NoTotal; i++) {
endPoints[i] = theNeurons[i - NoTotal]->F(x);
}
return endPoints;
}
double fullMGCRate::TransferInv(double in, double *params) {
double alpha = params[14];
double tr = params[16];
double lF;
lF = in / alpha / tr;
double out;
out = (lF - C) / M;
out = max(out, 0.0);
return out;
}
double fullMGCRate::TransferDeriv(double in, double *params) {
return (Transfer(in + 0.00005, params) - Transfer(in - 0.00005, params))
/ 0.0001;
}
double fullMGCRate::Transfer(double in, double *params) {
double alpha = params[14];
double tr = params[16];
double currentIn = max(in, 0.0);
double lF;
lF = M * currentIn + C;
lF = max(lF, 0.0);
return alpha * tr * lF;
}
#endif