use of umontreal.ssj.probdistmulti.BiNormalDist in project Stochastic-Inventory by RobinChen121.
the class Sampling method generateLHSamples.
/**
* latin hypercube sampling for binormal distribution.
*
* Since two independent variable, generate two variable independently, and merge the two samples into one
* @param distributions
* @param sampleNum
* @return a 2D random samples
*/
public double[][] generateLHSamples(BiNormalDist[] distributions, int sampleNum) {
int periodNum = distributions.length;
double[][] samples = new double[sampleNum][periodNum * 2];
double[][] samples1 = new double[sampleNum][periodNum];
double[][] samples2 = new double[sampleNum][periodNum];
// generate random possibility in [i/n, (i+1)/n], then get percent point function according to the possibility
for (int i = 0; i < periodNum; i++) {
NormalDist distribution1 = new NormalDist(distributions[i].getMu1(), distributions[i].getSigma1());
for (int j = 0; j < sampleNum; j++) {
// UniformGen.nextDouble(stream, 0, 1.0/sampleNum);
double randomNum = Math.random() / (double) sampleNum;
double lowBound = (double) j / (double) sampleNum;
samples1[j][i] = lowBound + randomNum;
samples1[j][i] = distribution1.inverseF(samples1[j][i]);
}
// ��������
shuffle(samples1);
}
for (int i = 0; i < periodNum; i++) {
NormalDist distribution2 = new NormalDist(distributions[i].getMu2(), distributions[i].getSigma2());
for (int j = 0; j < sampleNum; j++) {
// UniformGen.nextDouble(stream, 0, 1.0/sampleNum);
double randomNum = Math.random() / (double) sampleNum;
double lowBound = (double) j / (double) sampleNum;
samples2[j][i] = lowBound + randomNum;
samples2[j][i] = distribution2.inverseF(samples2[j][i]);
}
// ��������
shuffle(samples2);
}
for (int i = 0; i < sampleNum; i++) {
for (int j = 0; j < periodNum; j++) {
samples[i][j] = samples1[i][j];
samples[i][j + periodNum] = samples2[i][j];
}
}
return samples;
}
use of umontreal.ssj.probdistmulti.BiNormalDist in project Stochastic-Inventory by RobinChen121.
the class MultiItemCash method main.
public static void main(String[] args) {
double[] price = { 4, 50 };
// higher margin vs lower margin
double[] variCost = { 2, 4 };
// initial cash
double iniCash = 100;
// initial inventory
int iniInventory1 = 0;
int iniInventory2 = 0;
// higher average demand vs lower average demand
double[][] demand = { { 5, 6 }, { 5, 6 } };
// higher variance vs lower variance
double[] coe = { 0.25, 0.25 };
double[] salPrice = { 1, 1 };
// horizon length
int T = demand[0].length;
double truncationQuantile = 0.999;
int stepSize = 1;
double minCashState = 0;
double maxCashState = 10000;
int minInventoryState = 0;
int maxInventoryState = 200;
int Qbound = 100;
double discountFactor = 1;
// get demand possibilities for each period
BiNormalDist[] distributions = new BiNormalDist[T];
for (int t = 0; t < T; t++) distributions[t] = new BiNormalDist(demand[0][t], coe[0] * demand[0][t], demand[1][t], coe[1] * demand[1][t], 0);
// build action list for two items
Function<CashStateMulti, ArrayList<Actions>> buildActionList = s -> {
ArrayList<Actions> actions = new ArrayList<>();
for (int i = 0; i < Qbound; i++) for (int j = 0; j < Qbound; j++) {
if (variCost[0] * i + variCost[1] * j < s.getIniCash() + 0.1) {
Actions thisAction = new Actions(i, j);
actions.add(thisAction);
}
}
return actions;
};
// Immediate Value Function
ImmediateValueFunction<CashStateMulti, Actions, Demands, Double> immediateValue = (IniState, Actions, RandomDemands) -> {
double action1 = Actions.getFirstAction();
double action2 = Actions.getSecondAction();
double demand1 = RandomDemands.getFirstDemand();
double demand2 = RandomDemands.getSecondDemand();
double endInventory1 = Math.max(0, IniState.getIniInventory1() + action1 - demand1);
double endInventory2 = Math.max(0, IniState.getIniInventory2() + action2 - demand2);
double revenue1 = price[0] * (IniState.getIniInventory1() + action1 - endInventory1);
double revenue2 = price[1] * (IniState.getIniInventory2() + action2 - endInventory2);
double revenue = revenue1 + revenue2;
double orderingCost1 = variCost[0] * action1;
double orderingCost2 = variCost[1] * action2;
double orderingCosts = orderingCost1 + orderingCost2;
double salValue = 0;
if (IniState.getPeriod() == T) {
salValue = salPrice[0] * endInventory1 + salPrice[1] * endInventory2;
}
return revenue - orderingCosts + salValue;
};
// State Transition Function
StateTransitionFunction<CashStateMulti, Actions, Demands, CashStateMulti> stateTransition = (IniState, Actions, RandomDemands) -> {
double endInventory1 = IniState.getIniInventory1() + Actions.getFirstAction() - RandomDemands.getFirstDemand();
endInventory1 = Math.max(0, endInventory1);
double endInventory2 = IniState.getIniInventory2() + Actions.getSecondAction() - RandomDemands.getSecondDemand();
endInventory2 = Math.max(0, endInventory2);
double nextCash = IniState.getIniCash() + immediateValue.apply(IniState, Actions, RandomDemands);
nextCash = nextCash > maxCashState ? maxCashState : nextCash;
nextCash = nextCash < minCashState ? minCashState : nextCash;
endInventory1 = endInventory1 > maxInventoryState ? maxInventoryState : endInventory1;
endInventory2 = endInventory2 < minInventoryState ? minInventoryState : endInventory2;
// rounding states to save computing time
nextCash = (int) nextCash;
endInventory1 = (int) endInventory1;
endInventory2 = (int) endInventory2;
return new CashStateMulti(IniState.getPeriod() + 1, endInventory1, endInventory2, nextCash);
};
// GetPmfMulti pmfMulti = new GetPmfMulti(distributions, truncationQuantile, stepSize);
//
// /*******************************************************************
// * Solve
// */
// CashRecursionMulti recursion = new CashRecursionMulti(discountFactor, pmfMulti, buildActionList,
// stateTransition, immediateValue, T);
// int period = 1;
// CashStateMulti iniState = new CashStateMulti(period, iniInventory1, iniInventory2, iniCash);
// long currTime = System.currentTimeMillis();
// double finalValue = iniCash + recursion.getExpectedValue(iniState);
// System.out.println("final optimal cash is " + finalValue);
// System.out.println("optimal order quantity in the first priod is : Q1 = " + recursion.getAction(iniState).getFirstAction()
// + ", Q2 = " + recursion.getAction(iniState).getSecondAction());
// double time = (System.currentTimeMillis() - currTime) / 1000;
// System.out.println("running time is " + time + "s");
//
//
//
// /*******************************************************************
// * Simulating sdp results
// *
// * simulating results a little lower than SDP
// */
// int sampleNum = 10000;
// CashSimulationMulti simuation = new CashSimulationMulti(sampleNum, distributions, discountFactor,
// recursion, stateTransition, immediateValue);
// double simFinalValue = simuation.simulateSDPGivenSamplNum(iniState);
// System.out.println(simFinalValue);
/**
*****************************************************************
* try to find some ordering patters from optTable
*
* output results to excel
*/
// System.out.println("");
// double[][] optTable = recursion.getOptTable(variCost);
// WriteToExcel wr = new WriteToExcel();
// String fileName = "optTable" + "_c1=" + variCost[0] + "c2=" + variCost[1] + ".xls";
// String headString = "period" + "\t" + "x1" + "\t" + "x2" + "\t" + "w"+ "\t" + "R" + "\t" + "is limited cash and both ordering" + "\t" + "alpha"
// + "\t" + "Q1"+ "\t" + "Q2" + "\t" + "c1" + "\t" + "c2";
// wr.writeArrayToExcel(optTable, fileName, headString);
//
}
Aggregations