OptimizationProblem<LeastSquaresProblem.Evaluation>
LeastSquaresAdapter
public interface LeastSquaresProblem extends OptimizationProblem<LeastSquaresProblem.Evaluation>
Includes the observed values, computed model function, and
convergence/divergence criteria. Weights are implicit in LeastSquaresProblem.Evaluation.getResiduals()
and LeastSquaresProblem.Evaluation.getJacobian()
.
Instances are typically either created progressively using a builder
or created at once using a factory
.
Modifier and Type | Interface | Description |
---|---|---|
static interface |
LeastSquaresProblem.Evaluation |
An evaluation of a
LeastSquaresProblem at a particular point. |
Modifier and Type | Method | Description |
---|---|---|
LeastSquaresProblem.Evaluation |
evaluate(RealVector point) |
Evaluate the model at the specified point.
|
int |
getObservationSize() |
Get the number of observations (rows in the Jacobian) in this problem.
|
int |
getParameterSize() |
Get the number of parameters (columns in the Jacobian) in this problem.
|
RealVector |
getStart() |
Gets the initial guess.
|
getConvergenceChecker, getEvaluationCounter, getIterationCounter
RealVector getStart()
int getObservationSize()
int getParameterSize()
LeastSquaresProblem.Evaluation evaluate(RealVector point)
point
- the parameter values.MathIllegalStateException
- if the maximal number of evaluations (of the model vector function) is
exceeded.Copyright © 2016–2018 Hipparchus.org. All rights reserved.