use of uk.ac.sussex.gdsc.smlm.function.Gradient1Procedure in project GDSC-SMLM by aherbert.
the class UnivariateLikelihoodFisherInformationCalculator method compute.
/**
* {@inheritDoc}
*
* @throws DataException If the Fisher information cannot be computed for a function value
* @throws DataException If the Fisher information is infinite for a function value
*/
@Override
public FisherInformationMatrix compute(double[] parameters) {
final int n = gf.getNumberOfGradients();
final double[] data = new double[n * (n + 1) / 2];
gf.initialise1(parameters);
gf.forEach(new Gradient1Procedure() {
// CHECKSTYLE.OFF: MemberName
int k = -1;
// CHECKSTYLE.ON: MemberName
@Override
public void execute(double value, double[] dvDt) {
k++;
if (!fi[k].isValid(value)) {
return;
}
// Get the Fisher information of the value
final double f = fi[k].getFisherInformation(value);
if (f == 0) {
// No summation
return;
}
if (f == Double.POSITIVE_INFINITY) {
throw new DataException("Fisher information is infinite at f(" + k + ")");
}
// Compute the actual matrix data
for (int i = 0, c = 0; i < n; i++) {
final double wgt = f * dvDt[i];
for (int j = 0; j <= i; j++) {
data[c++] += wgt * dvDt[j];
}
}
}
});
// Generate symmetric matrix
final double[] matrix = new double[n * n];
for (int i = 0, c = 0; i < n; i++) {
for (int j = 0; j <= i; j++) {
matrix[i * n + j] = matrix[j * n + i] = data[c++];
}
}
return new FisherInformationMatrix(matrix, n);
}
use of uk.ac.sussex.gdsc.smlm.function.Gradient1Procedure in project GDSC-SMLM by aherbert.
the class ErfGaussian2DFunctionTest method functionComputesGradientForEachWith2Peaks.
@Test
void functionComputesGradientForEachWith2Peaks() {
Assumptions.assumeTrue(null != f2);
final ErfGaussian2DFunction f2 = (ErfGaussian2DFunction) this.f2;
final int n = f2.size();
final double[] du_da = new double[f2.getNumberOfGradients()];
final double[] du_db = new double[f2.getNumberOfGradients()];
final double[] d2u_da2 = new double[f2.getNumberOfGradients()];
final double[] values = new double[n];
final double[][] jacobian = new double[n][];
final double[][] jacobian2 = new double[n][];
for (final double background : testbackground) {
// Peak 1
for (final double signal1 : testsignal1) {
for (final double cx1 : testcx1) {
for (final double cy1 : testcy1) {
for (final double cz1 : testcz1) {
for (final double[] w1 : testw1) {
for (final double angle1 : testangle1) {
// Peak 2
for (final double signal2 : testsignal2) {
for (final double cx2 : testcx2) {
for (final double cy2 : testcy2) {
for (final double cz2 : testcz2) {
for (final double[] w2 : testw2) {
for (final double angle2 : testangle2) {
final double[] a = createParameters(background, signal1, cx1, cy1, cz1, w1[0], w1[1], angle1, signal2, cx2, cy2, cz2, w2[0], w2[1], angle2);
f2.initialiseExtended2(a);
// Compute single
for (int i = 0; i < n; i++) {
final double o1 = f2.eval(i, du_da);
final double o2 = f2.eval2(i, du_db, d2u_da2);
Assertions.assertEquals(o1, o2, 1e-10, "Value");
Assertions.assertArrayEquals(du_da, du_db, 1e-10, "Jacobian!=Jacobian");
values[i] = o1;
jacobian[i] = du_da.clone();
jacobian2[i] = d2u_da2.clone();
}
// Use procedures
f2.forEach(new ValueProcedure() {
int index = 0;
@Override
public void execute(double value) {
Assertions.assertEquals(values[index], value, 1e-10, "Value ValueProcedure");
index++;
}
});
f2.forEach(new Gradient1Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa) {
Assertions.assertEquals(values[index], value, 1e-10, "Value Gradient1Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da Gradient1Procedure");
index++;
}
});
f2.forEach(new Gradient2Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa, double[] d2yDa2) {
Assertions.assertEquals(values[index], value, 1e-10, "Value Gradient2Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da Gradient2Procedure");
Assertions.assertArrayEquals(jacobian2[index], d2yDa2, 1e-10, "d2u_da2 Gradient2Procedure");
index++;
}
});
f2.forEach(new ExtendedGradient2Procedure() {
int index = 0;
@Override
public void executeExtended(double value, double[] dyDa, double[] d2yDaDb) {
Assertions.assertEquals(values[index], value, 1e-10, "Value ExtendedGradient2Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da ExtendedGradient2Procedure");
for (int j = 0, k = 0; j < d2u_da2.length; j++, k += d2u_da2.length + 1) {
d2u_da2[j] = d2yDaDb[k];
}
Assertions.assertArrayEquals(jacobian2[index], d2u_da2, 1e-10, "d2u_da2 Gradient2Procedure");
index++;
}
});
}
}
}
}
}
}
}
}
}
}
}
}
}
}
use of uk.ac.sussex.gdsc.smlm.function.Gradient1Procedure in project GDSC-SMLM by aherbert.
the class Gaussian2DFunction method computeValuesAndJacobian.
@Override
public Pair<double[], double[][]> computeValuesAndJacobian(double[] variables) {
initialise1(variables);
final int n = size();
final double[][] jacobian = new double[n][];
final double[] values = new double[n];
forEach(new Gradient1Procedure() {
int index;
@Override
public void execute(double value, double[] derivative) {
values[index] = value;
jacobian[index++] = derivative.clone();
}
});
return Pair.of(values, jacobian);
}
use of uk.ac.sussex.gdsc.smlm.function.Gradient1Procedure in project GDSC-SMLM by aherbert.
the class ErfGaussian2DFunctionTest method functionComputesGradientForEach.
@Test
void functionComputesGradientForEach() {
final ErfGaussian2DFunction f1 = (ErfGaussian2DFunction) this.f1;
final int n = f1.size();
final double[] du_da = new double[f1.getNumberOfGradients()];
final double[] du_db = new double[f1.getNumberOfGradients()];
final double[] d2u_da2 = new double[f1.getNumberOfGradients()];
final double[] values = new double[n];
final double[][] jacobian = new double[n][];
final double[][] jacobian2 = new double[n][];
for (final double background : testbackground) {
// Peak 1
for (final double signal1 : testsignal1) {
for (final double cx1 : testcx1) {
for (final double cy1 : testcy1) {
for (final double cz1 : testcz1) {
for (final double[] w1 : testw1) {
for (final double angle1 : testangle1) {
final double[] a = createParameters(background, signal1, cx1, cy1, cz1, w1[0], w1[1], angle1);
f1.initialiseExtended2(a);
// Compute single
for (int i = 0; i < n; i++) {
final double o1 = f1.eval(i, du_da);
final double o2 = f1.eval2(i, du_db, d2u_da2);
Assertions.assertEquals(o1, o2, 1e-10, "Value");
Assertions.assertArrayEquals(du_da, du_db, 1e-10, "Jacobian!=Jacobian");
values[i] = o1;
jacobian[i] = du_da.clone();
jacobian2[i] = d2u_da2.clone();
}
// Use procedures
f1.forEach(new ValueProcedure() {
int index = 0;
@Override
public void execute(double value) {
Assertions.assertEquals(values[index], value, 1e-10, "Value ValueProcedure");
index++;
}
});
f1.forEach(new Gradient1Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa) {
Assertions.assertEquals(values[index], value, 1e-10, "Value Gradient1Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da Gradient1Procedure");
index++;
}
});
f1.forEach(new Gradient2Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa, double[] d2yDa2) {
Assertions.assertEquals(values[index], value, 1e-10, "Value Gradient2Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da Gradient2Procedure");
Assertions.assertArrayEquals(jacobian2[index], d2yDa2, 1e-10, "d2u_da2 Gradient2Procedure");
index++;
}
});
f1.forEach(new ExtendedGradient2Procedure() {
int index = 0;
@Override
public void executeExtended(double value, double[] dyDa, double[] d2yDaDb) {
Assertions.assertEquals(values[index], value, 1e-10, "Value ExtendedGradient2Procedure");
Assertions.assertArrayEquals(jacobian[index], dyDa, 1e-10, "du_da ExtendedGradient2Procedure");
for (int j = 0, k = 0; j < d2u_da2.length; j++, k += d2u_da2.length + 1) {
d2u_da2[j] = d2yDaDb[k];
}
Assertions.assertArrayEquals(jacobian2[index], d2u_da2, 1e-10, "d2u_da2 Gradient2Procedure");
index++;
}
});
}
}
}
}
}
}
}
}
use of uk.ac.sussex.gdsc.smlm.function.Gradient1Procedure in project GDSC-SMLM by aherbert.
the class PsfModelGradient1FunctionTest method canComputeValueAndGradient.
@Test
void canComputeValueAndGradient() {
// Use a reasonable z-depth function from the Smith, et al (2010) paper (page 377)
final double sx = 1.08;
final double sy = 1.01;
final double gamma = 0.389;
final double d = 0.531;
final double Ax = -0.0708;
final double Bx = -0.073;
final double Ay = 0.164;
final double By = 0.0417;
final AstigmatismZModel zModel = HoltzerAstigmatismZModel.create(sx, sy, gamma, d, Ax, Bx, Ay, By);
// Small size ensure the PSF model covers the entire image
final int maxx = 11;
final int maxy = 11;
final double[] ve = new double[maxx * maxy];
final double[] vo = new double[maxx * maxy];
final double[][] ge = new double[maxx * maxy][];
final double[][] go = new double[maxx * maxy][];
final PsfModelGradient1Function psf = new PsfModelGradient1Function(new GaussianPsfModel(zModel), maxx, maxy);
final ErfGaussian2DFunction f = new SingleAstigmatismErfGaussian2DFunction(maxx, maxy, zModel);
f.setErfFunction(ErfFunction.COMMONS_MATH);
final double[] a2 = new double[Gaussian2DFunction.PARAMETERS_PER_PEAK + 1];
final DoubleDoubleBiPredicate equality = TestHelper.doublesAreClose(1e-8, 0);
final double c = maxx * 0.5;
for (int i = -1; i <= 1; i++) {
final double x0 = c + i * 0.33;
for (int j = -1; j <= 1; j++) {
final double x1 = c + j * 0.33;
for (int k = -1; k <= 1; k++) {
final double x2 = k * 0.33;
for (final double in : new double[] { 23.2, 405.67 }) {
// Background is constant for gradients so just use 1 value
final double[] a = new double[] { 2.2, in, x0, x1, x2 };
psf.initialise1(a);
psf.forEach(new Gradient1Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa) {
vo[index] = value;
go[index] = dyDa.clone();
index++;
}
});
a2[Gaussian2DFunction.BACKGROUND] = a[0];
a2[Gaussian2DFunction.SIGNAL] = a[1];
a2[Gaussian2DFunction.X_POSITION] = a[2] - 0.5;
a2[Gaussian2DFunction.Y_POSITION] = a[3] - 0.5;
a2[Gaussian2DFunction.Z_POSITION] = a[4];
f.initialise1(a2);
f.forEach(new Gradient1Procedure() {
int index = 0;
@Override
public void execute(double value, double[] dyDa) {
ve[index] = value;
ge[index] = dyDa.clone();
index++;
}
});
for (int ii = 0; ii < ve.length; ii++) {
TestAssertions.assertTest(ve[ii], vo[ii], equality);
TestAssertions.assertArrayTest(ge[ii], go[ii], equality);
}
}
}
}
}
}
Aggregations