/**
* <p>
* Solves for x in the following equation:<br>
* <br>
* A*x = b
* </p>
*
* <p>
* If the system could not be solved then false is returned. If it returns true
* that just means the algorithm finished operating, but the results could still be bad
* because 'A' is singular or nearly singular.
* </p>
*
* <p>
* If repeat calls to solve are being made then one should consider using {@link LinearSolverFactory_DSCC}
* instead.
* </p>
*
* <p>
* It is ok for 'b' and 'x' to be the same matrix.
* </p>
*
* @param a (Input) A matrix that is m by n. Not modified.
* @param b (Input) A matrix that is n by k. Not modified.
* @param x (Output) A matrix that is m by k. Modified.
*
* @return true if it could invert the matrix false if it could not.
*/
public static bool solve(DMatrixSparseCSC a,
DMatrixRMaj b,
DMatrixRMaj x)
{
LinearSolverSparse <DMatrixSparseCSC, DMatrixRMaj> solver;
if (a.numRows > a.numCols)
{
solver = LinearSolverFactory_DSCC.qr(FillReducing.NONE); // todo specify a filling that makes sense
}
else
{
solver = LinearSolverFactory_DSCC.lu(FillReducing.NONE);
}
// Ensure that the input isn't modified
if (solver.modifiesA())
{
a = (DMatrixSparseCSC)a.copy();
}
if (solver.modifiesB())
{
b = (DMatrixRMaj)b.copy();
}
// decompose then solve the matrix
if (!solver.setA(a))
{
return(false);
}
solver.solve(b, x);
return(true);
}
/**
* <p>
* Checks to see if the matrix is positive semidefinite:
* </p>
* <p>
* x<sup>T</sup> A x ≥ 0<br>
* for all x where x is a non-zero vector and A is a symmetric matrix.
* </p>
*
* @param A square symmetric matrix. Not modified.
*
* @return True if it is positive semidefinite and false if it is not.
*/
public static bool isPositiveSemidefinite(DMatrixRMaj A)
{
if (!isSquare(A))
{
return(false);
}
EigenDecomposition_F64 <DMatrixRMaj> eig = DecompositionFactory_DDRM.eig(A.numCols, false);
if (eig.inputModified())
{
A = (DMatrixRMaj)A.copy();
}
eig.decompose(A);
for (int i = 0; i < A.numRows; i++)
{
Complex_F64 v = eig.getEigenvalue(i);
if (v.getReal() < 0)
{
return(false);
}
}
return(true);
}
/**
* Computes the QR decomposition of the provided matrix.
*
* @param A Matrix which is to be decomposed. Not modified.
*/
public void decompose(DMatrixRMaj A)
{
Equation.Equation eq = new Equation.Equation();
this.QR = (DMatrixRMaj)A.copy();
int N = Math.Min(A.numCols, A.numRows);
gammas = new double[A.numCols];
for (int i = 0; i < N; i++)
{
// update temporary variables
eq.alias(QR.numRows - i, "Ni", QR, "QR", i, "i");
// Place the column that should be zeroed into v
eq.process("v=QR(i:,i)");
eq.process("maxV=max(abs(v))");
// Note that v is lazily created above. Need direct access to it, which is done below.
DMatrixRMaj v = eq.lookupMatrix("v");
double maxV = eq.lookupDouble("maxV");
if (maxV > 0 && v.getNumElements() > 1)
{
// normalize to reduce overflow issues
eq.process("v=v/maxV");
// compute the magnitude of the vector
double tau = NormOps_DDRM.normF(v);
if (v.get(0) < 0)
{
tau *= -1.0;
}
eq.alias(tau, "tau");
eq.process("u_0 = v(0,0)+tau");
eq.process("gamma = u_0/tau");
eq.process("v=v/u_0");
eq.process("v(0,0)=1");
eq.process("QR(i:,i:) = (eye(Ni) - gamma*v*v')*QR(i:,i:)");
eq.process("QR(i:,i) = v");
eq.process("QR(i,i) = -1*tau*maxV");
// save gamma for recomputing Q later on
gammas[i] = eq.lookupDouble("gamma");
}
}
}
/**
* <p>
* Checks to see if the matrix is positive definite.
* </p>
* <p>
* x<sup>T</sup> A x > 0<br>
* for all x where x is a non-zero vector and A is a symmetric matrix.
* </p>
*
* @param A square symmetric matrix. Not modified.
* @return True if it is positive definite and false if it is not.
*/
public static bool isPositiveDefinite(DMatrixRMaj A)
{
if (!isSquare(A))
{
return(false);
}
CholeskyDecompositionInner_DDRM chol = new CholeskyDecompositionInner_DDRM(true);
if (chol.inputModified())
{
A = A.copy();
}
return(chol.decompose(A));
}
/**
* Creates a random distribution with the specified mean and covariance. The references
* to the variables are not saved, their value are copied.
*
* @param rand Used to create the random numbers for the draw. Reference is saved.
* @param cov The covariance of the distribution. Not modified.
*/
public CovarianceRandomDraw_DDRM(IMersenneTwister rand, DMatrixRMaj cov)
{
r = new DMatrixRMaj(cov.numRows, 1);
CholeskyDecompositionInner_DDRM cholesky = new CholeskyDecompositionInner_DDRM(true);
if (cholesky.inputModified())
{
cov = (DMatrixRMaj)cov.copy();
}
if (!cholesky.decompose(cov))
{
throw new InvalidOperationException("Decomposition failed!");
}
A = cholesky.getT();
this.rand = rand;
}
/**
* Computes the nullity of a matrix using the specified tolerance.
*
* @param A Matrix whose rank is to be calculated. Not modified.
* @param threshold The numerical threshold used to determine a singular value.
* @return The matrix's nullity.
*/
public static int nullity(DMatrixRMaj A, double threshold)
{
SingularValueDecomposition_F64 <DMatrixRMaj> svd =
DecompositionFactory_DDRM.svd(A.numRows, A.numCols, false, false, true);
if (svd.inputModified())
{
A = (DMatrixRMaj)A.copy();
}
if (!svd.decompose(A))
{
throw new InvalidOperationException("Decomposition failed");
}
return(SingularOps_DDRM.nullity(svd, threshold));
}
/**
* Checks to see if the rows of the provided matrix are linearly independent.
*
* @param A Matrix whose rows are being tested for linear independence.
* @return true if linearly independent and false otherwise.
*/
public static bool isRowsLinearIndependent(DMatrixRMaj A)
{
// LU decomposition
LUDecomposition <DMatrixRMaj> lu = DecompositionFactory_DDRM.lu(A.numRows, A.numCols);
if (lu.inputModified())
{
A = A.copy();
}
if (!lu.decompose(A))
{
throw new SystemException("Decompositon failed?");
}
// if they are linearly independent it should not be singular
return(!lu.isSingular());
}
/**
* Creates a random distribution with the specified mean and covariance. The references
* to the variables are not saved, their value are copied.
*
* @param rand Used to create the random numbers for the draw. Reference is saved.
* @param cov The covariance of the distribution. Not modified.
*/
public CovarianceRandomDraw_DDRM(Random rand, DMatrixRMaj cov)
{
r = new DMatrixRMaj(cov.numRows, 1);
CholeskyDecompositionInner_DDRM cholesky = new CholeskyDecompositionInner_DDRM(true);
if (cholesky.inputModified())
{
cov = cov.copy();
}
if (!cholesky.decompose(cov))
{
throw new SystemException("Decomposition failed!");
}
A = cholesky.getT();
//this.rand = rand;
this.rand = new Java.Util.Random();
}
public override object Clone()
{
CMAESSpecies myobj = (CMAESSpecies)(base.Clone());
// clone the distribution and other variables here
myobj.c = c.copy();
myobj.b = b.copy();
myobj.d = d.copy();
myobj.bd = (DMatrixRMaj)bd.copy();
myobj.sbd = (DMatrixRMaj)sbd.copy();
myobj.invsqrtC = invsqrtC.copy();
myobj.xmean = xmean.copy();
myobj.ps = ps.copy();
myobj.pc = pc.copy();
return(myobj);
}
public void setup(DMatrixRMaj A)
{
if (A.numRows != A.numCols)
{
throw new InvalidOperationException("Must be square");
}
if (N != A.numRows)
{
N = A.numRows;
this.A = (DMatrixRMaj)A.copy();
u = new DMatrixRMaj(A.numRows, 1);
_temp = new DMatrixRMaj(A.numRows, 1);
numStepsFind = new int[A.numRows];
}
else
{
this.A.set(A);
UtilEjml.memset(numStepsFind, 0, numStepsFind.Length);
}
// zero all the off numbers that should be zero for a hessenberg matrix
for (int i = 2; i < N; i++)
{
for (int j = 0; j < i - 1; j++)
{
this.A.set(i, j, 0);
}
}
eigenvalues = new Complex_F64[A.numRows];
for (int i = 0; i < eigenvalues.Length; i++)
{
eigenvalues[i] = new Complex_F64();
}
numEigen = 0;
lastExceptional = 0;
numExceptional = 0;
steps = 0;
}
private List <DMatrixRMaj> createSimulatedMeas(DMatrixRMaj x)
{
List <DMatrixRMaj> ret = new List <DMatrixRMaj>();
DMatrixRMaj F = createF(T);
DMatrixRMaj H = createH();
// UtilEjml.print(F);
// UtilEjml.print(H);
DMatrixRMaj x_next = new DMatrixRMaj(x);
DMatrixRMaj z = new DMatrixRMaj(H.numRows, 1);
for (int i = 0; i < MAX_STEPS; i++)
{
CommonOps_DDRM.mult(F, x, x_next);
CommonOps_DDRM.mult(H, x_next, z);
ret.Add((DMatrixRMaj)z.copy());
x.set(x_next);
}
return(ret);
}
/**
* Computes the QR decomposition of the provided matrix.
*
* @param A Matrix which is to be decomposed. Not modified.
*/
public void decompose(DMatrixRMaj A)
{
this.QR = (DMatrixRMaj)A.copy();
int N = Math.Min(A.numCols, A.numRows);
gammas = new double[A.numCols];
DMatrixRMaj A_small = new DMatrixRMaj(A.numRows, A.numCols);
DMatrixRMaj A_mod = new DMatrixRMaj(A.numRows, A.numCols);
DMatrixRMaj v = new DMatrixRMaj(A.numRows, 1);
DMatrixRMaj Q_k = new DMatrixRMaj(A.numRows, A.numRows);
for (int i = 0; i < N; i++)
{
// reshape temporary variables
A_small.reshape(QR.numRows - i, QR.numCols - i, false);
A_mod.reshape(A_small.numRows, A_small.numCols, false);
v.reshape(A_small.numRows, 1, false);
Q_k.reshape(v.getNumElements(), v.getNumElements(), false);
// use extract matrix to get the column that is to be zeroed
CommonOps_DDRM.extract(QR, i, QR.numRows, i, i + 1, v, 0, 0);
double max = CommonOps_DDRM.elementMaxAbs(v);
if (max > 0 && v.getNumElements() > 1)
{
// normalize to reduce overflow issues
CommonOps_DDRM.divide(v, max);
// compute the magnitude of the vector
double tau = NormOps_DDRM.normF(v);
if (v.get(0) < 0)
{
tau *= -1.0;
}
double u_0 = v.get(0) + tau;
double gamma = u_0 / tau;
CommonOps_DDRM.divide(v, u_0);
v.set(0, 1.0);
// extract the submatrix of A which is being operated on
CommonOps_DDRM.extract(QR, i, QR.numRows, i, QR.numCols, A_small, 0, 0);
// A = (I - γ*u*u<sup>T</sup>)A
CommonOps_DDRM.setIdentity(Q_k);
CommonOps_DDRM.multAddTransB(-gamma, v, v, Q_k);
CommonOps_DDRM.mult(Q_k, A_small, A_mod);
// save the results
CommonOps_DDRM.insert(A_mod, QR, i, i);
CommonOps_DDRM.insert(v, QR, i, i);
QR.unsafe_set(i, i, -tau * max);
// save gamma for recomputing Q later on
gammas[i] = gamma;
}
}
}
