Examples with NumberContext org.ojalgo.type.context.NumberContext used on opensource projects

Example 16 with NumberContext

use of org.ojalgo.type.context.NumberContext in project ojAlgo by optimatika.

the class ConvexProblems method testP20150809.

/**
 * Issue reported at GitHub. A set of problems related to when Q is zero - a linear problem. Generally the
 * ConvexSolver is not the right option to handle linear problems, but there is some desireable behaviour.
 */
@Test
public void testP20150809() {
    final NumberContext precision = new NumberContext(11, 14, RoundingMode.HALF_EVEN);
    final Primitive64Array tmpExpectedSolution = Primitive64Array.wrap(new double[] { 0.12, -0.05, 0.08, 0.07 });
    final Primitive64Array tmpBoundedSolution = Primitive64Array.wrap(new double[] { 99999, -99999, 99999, 99999 });
    ConvexSolver tmpSolver = P20150809.buildModel(true, false);
    Result tmpResult = tmpSolver.solve();
    TestUtils.assertStateNotLessThanOptimal(tmpResult);
    TestUtils.assertEquals(tmpExpectedSolution, tmpResult, precision);
    tmpSolver = P20150809.buildModel(true, true);
    tmpResult = tmpSolver.solve();
    TestUtils.assertStateNotLessThanOptimal(tmpResult);
    TestUtils.assertEquals(tmpExpectedSolution, tmpResult, precision);
    tmpSolver = P20150809.buildModel(false, false);
    tmpResult = tmpSolver.solve();
    TestUtils.assertEquals(Optimisation.State.UNBOUNDED, tmpResult.getState());
    tmpSolver = P20150809.buildModel(false, true);
    tmpResult = tmpSolver.solve();
    // Since it is now constrained, the solver should be able find the optimal solution.
    TestUtils.assertStateNotLessThanOptimal(tmpResult);
    TestUtils.assertEquals(tmpBoundedSolution, tmpResult, precision);
}

Example 17 with NumberContext

use of org.ojalgo.type.context.NumberContext in project ojAlgo by optimatika.

the class ConvexProblems method testP20111205.

/**
 * Continuation of P20111129 http://bugzilla.optimatika.se/show_bug.cgi?id=15 Have to turn off validation
 * as Q is not positive semidefinite.
 * <p>
 * 2016-03-07: Initially the solution (from AMPL/LOQO) was stated to be:
 * <p>
 *
 * <pre>
 * 1.78684, 0.000326128, 1.78665, 0.000136478, 495.429, 0.00358488, 495.427, 0.00178874, 8.90701, 0.000339811, 8.90684, 0.000174032
 * </pre>
 * <p>
 * The ExpressionsBasedModel can only validate this solution to be correct using a very "poor" accuracy
 * context. When ExpressionsBasedModel uses CPLEX as the solver a slightly different solution is returned
 * that validates much better. Switched to using that solution as the expected solution in this test:
 * <p>
 *
 * <pre>
 * 1.7856570552, 1.216415374E-5, 1.78565097263, 6.08157995E-6, 495.426247828, 2.478968927E-5, 495.426235433, 1.239483719E-5, 8.90673094088, 6.04347562E-6, 8.90672791911, 3.02171321E-6
 * </pre>
 * </p>
 */
@Test
public void testP20111205() {
    final PrimitiveDenseStore tmpAE = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, -1.0, -1.0, 1.0 }, { 1.0, -1.0, -1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 1.0, -1.0, -1.0, 1.0, 0.0, 0.0, 0.0, 0.0 } });
    final PrimitiveDenseStore tmpBE = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 0.0 }, { 0.0 }, { 0.0 } });
    final PrimitiveDenseStore tmpQ = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 42.58191012032541, -42.58191012032541, 0.0, 0.0, 0.029666091804595635, -0.029666091804595635, 0.0, 0.0, 9.954580659495097, -9.954580659495097, 0.0, 0.0 }, { -42.58191012032541, 42.58191012032541, 0.0, 0.0, -0.029666091804595635, 0.029666091804595635, 0.0, 0.0, -9.954580659495097, 9.954580659495097, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.029666091804595635, -0.029666091804595635, 0.0, 0.0, 0.8774199042430086, -0.8774199042430086, 0.0, 0.0, -3.537087573378497, 3.537087573378497, 0.0, 0.0 }, { -0.029666091804595635, 0.029666091804595635, 0.0, 0.0, -0.8774199042430086, 0.8774199042430086, 0.0, 0.0, 3.537087573378497, -3.537087573378497, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 9.954580659495097, -9.954580659495097, 0.0, 0.0, -3.537087573378497, 3.537087573378497, 0.0, 0.0, 153.76101274121527, -153.76101274121527, 0.0, 0.0 }, { -9.954580659495097, 9.954580659495097, 0.0, 0.0, 3.537087573378497, -3.537087573378497, 0.0, 0.0, -153.76101274121527, 153.76101274121527, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 } });
    final PrimitiveDenseStore tmpC = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 185.8491751747291 }, { -192.3021967647291 }, { -6.45302159 }, { -6.45302159 }, { 406.4118818820076 }, { -409.5778277520076 }, { -3.16594587 }, { -3.16594587 }, { -352.0970015985486 }, { 339.11043506854867 }, { -12.986566530000001 }, { -12.986566530000001 } });
    final PrimitiveDenseStore tmpAI = PrimitiveDenseStore.FACTORY.rows(new double[][] { { -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0, 0.0 }, { 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, -1.0 } });
    final PrimitiveDenseStore tmpBI = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 }, { 0.0 } });
    // Optimisation.Result from CPLEX (via ExpressionsBasedModel)
    final double[] tmpExpected = new double[] { 1.7856570552, 1.216415374E-5, 1.78565097263, 6.08157995E-6, 495.426247828, 2.478968927E-5, 495.426235433, 1.239483719E-5, 8.90673094088, 6.04347562E-6, 8.90672791911, 3.02171321E-6 };
    final PrimitiveDenseStore[] tmpMatrices = new PrimitiveDenseStore[6];
    tmpMatrices[0] = tmpAE;
    tmpMatrices[1] = tmpBE;
    tmpMatrices[2] = tmpQ;
    tmpMatrices[3] = tmpC;
    tmpMatrices[4] = tmpAI;
    tmpMatrices[5] = tmpBI;
    // The original AMPL/LOQO solution was given with 6 digits precision and never more than 9 decimals
    // ojAlgo can only get roughly the same solution
    final NumberContext tmpAccuracy = NumberContext.getGeneral(3, 3);
    ConvexProblems.builAndTestModel(tmpMatrices, tmpExpected, tmpAccuracy, true);
}

Example 18 with NumberContext

use of org.ojalgo.type.context.NumberContext in project ojAlgo by optimatika.

the class ConvexProblems method doEarly2008.

static void doEarly2008(final Variable[] variables, final Access2D<?> covariances, final Access1D<?> expected) {
    final ExpressionsBasedModel tmpModel = new ExpressionsBasedModel(variables);
    final Expression tmpVariance = tmpModel.addExpression("Variance");
    tmpVariance.setQuadraticFactors(tmpModel.getVariables(), covariances);
    tmpVariance.weight(BigMath.PI.multiply(BigMath.E).multiply(BigMath.HALF));
    final Expression tmpBalance = tmpModel.addExpression("Balance");
    tmpBalance.setLinearFactorsSimple(tmpModel.getVariables());
    tmpBalance.level(BigMath.ONE);
    // tmpModel.options.debug(ConvexSolver.class);
    // tmpModel.options.validate = false;
    final Result tmpActualResult = tmpModel.minimise();
    final NumberContext tmpAccuracy = StandardType.PERCENT.newPrecision(5);
    TestUtils.assertTrue(tmpModel.validate(Array1D.BIG.copy(expected), tmpAccuracy));
    TestUtils.assertTrue(tmpModel.validate(tmpActualResult, tmpAccuracy));
    final TwiceDifferentiable<Double> tmpObjective = tmpModel.objective().toFunction();
    final double tmpExpObjFuncVal = tmpObjective.invoke(Access1D.asPrimitive1D(expected));
    final double tmpActObjFuncVal = tmpObjective.invoke(Access1D.asPrimitive1D(tmpActualResult));
    TestUtils.assertEquals(tmpExpObjFuncVal, tmpActObjFuncVal, tmpAccuracy);
    TestUtils.assertEquals(expected, tmpActualResult, tmpAccuracy);
    // Test that the LinearSolver can determine feasibility
    final ExpressionsBasedModel relaxedModel = tmpModel.relax(false);
    final Optimisation.Result tmpLinearResult = relaxedModel.minimise();
    TestUtils.assertStateNotLessThanFeasible(tmpLinearResult);
    OptimisationConvexTests.assertDirectAndIterativeEquals(tmpModel, tmpAccuracy);
}

Example 19 with NumberContext

use of org.ojalgo.type.context.NumberContext in project ojAlgo by optimatika.

the class ConvexProblems method testP20111129.

/**
 * ojAlgo could not solve this, but LOQO (and others) could. I re-implemented the problem code just to
 * verify there was no problem there. http://bugzilla.optimatika.se/show_bug.cgi?id=11
 */
@Test
public void testP20111129() {
    final Variable x1 = new Variable("X1");
    final Variable x2 = new Variable("X2").lower(BigMath.HUNDRED.negate()).upper(BigMath.HUNDRED);
    final Variable x3 = new Variable("X3").lower(BigMath.ZERO);
    final Variable x4 = new Variable("X4").lower(BigMath.ZERO);
    final Variable[] tmpVariables = new Variable[] { x1, x2, x3, x4 };
    final ExpressionsBasedModel tmpModel = new ExpressionsBasedModel(tmpVariables);
    final Expression tmpObjExpr = tmpModel.addExpression("Objective");
    tmpModel.setMinimisation();
    tmpObjExpr.set(2, 2, BigMath.HALF);
    tmpObjExpr.set(3, 3, BigMath.TWO);
    tmpObjExpr.set(2, 3, BigMath.TWO.negate());
    tmpObjExpr.set(0, BigMath.THREE);
    tmpObjExpr.set(1, BigMath.TWO.negate());
    tmpObjExpr.set(2, BigMath.ONE);
    tmpObjExpr.set(3, BigMath.FOUR.negate());
    tmpObjExpr.weight(BigMath.ONE);
    Expression tmpConstrExpr;
    tmpConstrExpr = tmpModel.addExpression("C1").lower(BigMath.FOUR);
    tmpConstrExpr.set(0, BigMath.ONE);
    tmpConstrExpr.set(1, BigMath.ONE);
    tmpConstrExpr.set(2, BigMath.FOUR.negate());
    tmpConstrExpr.set(3, BigMath.TWO);
    tmpConstrExpr = tmpModel.addExpression("C2").upper(BigMath.SIX);
    tmpConstrExpr.set(0, BigMath.THREE.negate());
    tmpConstrExpr.set(1, BigMath.ONE);
    tmpConstrExpr.set(2, BigMath.TWO.negate());
    tmpConstrExpr = tmpModel.addExpression("C3").level(BigMath.NEG);
    tmpConstrExpr.set(1, BigMath.ONE);
    tmpConstrExpr.set(3, BigMath.NEG);
    tmpConstrExpr = tmpModel.addExpression("C4").level(BigMath.ZERO);
    tmpConstrExpr.set(0, BigMath.ONE);
    tmpConstrExpr.set(1, BigMath.ONE);
    tmpConstrExpr.set(2, BigMath.NEG);
    // tmpModel.options.debug(ConvexSolver.class);
    final Result tmpResult = tmpModel.minimise();
    final double tmpObjFuncVal = tmpResult.getValue();
    TestUtils.assertEquals(-5.281249989, tmpObjFuncVal, new NumberContext(7, 6));
    final double[] tmpExpected = new double[] { -1.1875, 1.5625, 0.375, 2.5625 };
    for (int i = 0; i < tmpExpected.length; i++) {
        TestUtils.assertEquals(tmpExpected[i], tmpVariables[i].getValue().doubleValue(), new NumberContext(5, 4));
    }
    OptimisationConvexTests.assertDirectAndIterativeEquals(tmpModel, null);
}

Example 20 with NumberContext

use of org.ojalgo.type.context.NumberContext in project ojAlgo by optimatika.

the class LinearDesignTestCases method test4LinearModelCase.

/**
 * http://www.saintmarys.edu/~psmith/338act14.html
 */
@Test
public void test4LinearModelCase() {
    final Variable[] tmpVariables = new Variable[] { new Variable("X1").lower(ZERO).weight(ONE.negate()), new Variable("X2").lower(ZERO).weight(ONE), new Variable("X3").lower(ZERO).weight(ZERO) };
    final ExpressionsBasedModel tmpModel = new ExpressionsBasedModel(tmpVariables);
    final Expression tmpExprC1 = tmpModel.addExpression("C1");
    for (int i = 0; i < tmpModel.countVariables(); i++) {
        tmpExprC1.set(i, new BigDecimal[] { SIX, ONE.negate(), ZERO }[i]);
    }
    tmpExprC1.upper(TEN);
    final Expression tmpExprC2 = tmpModel.addExpression("C2");
    for (int i = 0; i < tmpModel.countVariables(); i++) {
        tmpExprC2.set(i, new BigDecimal[] { ONE, FIVE, ZERO }[i]);
    }
    tmpExprC2.lower(FOUR);
    final Expression tmpExprC3 = tmpModel.addExpression("C3");
    for (int i = 0; i < tmpModel.countVariables(); i++) {
        tmpExprC3.set(i, new BigDecimal[] { ONE, FIVE, ONE }[i]);
    }
    tmpExprC3.level(FIVE);
    final Optimisation.Result tmpResult = tmpModel.minimise();
    final BasicMatrix tmpSolution = RationalMatrix.FACTORY.columns(tmpResult);
    final PhysicalStore<Double> tmpExpX = PrimitiveDenseStore.FACTORY.rows(new double[][] { { 1.74 }, { 0.45 }, { 1.0 } });
    final PhysicalStore<Double> tmpActX = PrimitiveDenseStore.FACTORY.copy(tmpSolution.selectRows(new int[] { 0, 1, 2 }));
    tmpActX.modifyAll(new NumberContext(7, 2).getFunction(PrimitiveFunction.getSet()));
    TestUtils.assertEquals(tmpExpX, tmpActX);
}