kmath-optimization/space.kscience.kmath.optimization

Package space.kscience.kmath.optimization

Types

FunctionOptimization

class FunctionOptimization<T>(features: FeatureSet<OptimizationFeature>, expression: DifferentiableExpression<T>) : OptimizationProblem<T>

FunctionOptimizationBuilder

class FunctionOptimizationBuilder<T>(expression: DifferentiableExpression<T>) : OptimizationBuilder<T, FunctionOptimization<T>>

FunctionOptimizationTarget

enum FunctionOptimizationTarget : Enum<FunctionOptimizationTarget> , OptimizationFeature

OptimizationBuilder

abstract class OptimizationBuilder<T, R : OptimizationProblem<T>>

OptimizationCovariance

class OptimizationCovariance<T>(covariance: Matrix<T>) : OptimizationFeature

OptimizationFeature

interface OptimizationFeature : Feature<OptimizationFeature>

OptimizationLog

class OptimizationLog(loggable: Loggable) : Loggable, OptimizationFeature

OptimizationParameters

class OptimizationParameters(symbols: List<Symbol>) : OptimizationFeature

OptimizationPrior

interface OptimizationPrior<T> : OptimizationFeature, DifferentiableExpression<T>

OptimizationProblem

interface OptimizationProblem<T> : Featured<OptimizationFeature>

OptimizationResult

open class OptimizationResult<T>(point: Map<Symbol, T>) : OptimizationFeature

OptimizationStartPoint

open class OptimizationStartPoint<T>(point: Map<Symbol, T>) : OptimizationFeature

OptimizationValue

class OptimizationValue<T>(value: T) : OptimizationFeature

interface Optimizer<T, P : OptimizationProblem<T>>

PointToCurveDistance

interface PointToCurveDistance : OptimizationFeature

Specify the way to compute distance from point to the curve as DifferentiableExpression

interface PointWeight : OptimizationFeature

Compute a wight of the point. The more the weight, the more impact this point will have on the fit. By default, uses Dispersion^-1

@UnstableKMathAPI

object QowOptimizer : Optimizer<Double, XYFit>

An optimizer based onf Fyodor Tkachev's quasi-optimal weights method. See the article.

class QowRuns(runs: Int) : OptimizationFeature

class XYFit(data: XYColumnarData<Double, Double, Double>, model: DifferentiableExpression<Double>, features: FeatureSet<OptimizationFeature>, pointToCurveDistance: PointToCurveDistance, pointWeight: PointWeight, xSymbol: Symbol) : OptimizationProblem<Double>

A fit problem for X-Y-Yerr data. Also known as "least-squares" problem.

XYOptimizationBuilder

class XYOptimizationBuilder(data: XYColumnarData<Double, Double, Double>, model: DifferentiableExpression<Double>) : OptimizationBuilder<Double, XYFit>

Functions

suspend fun <I : Any, A : ExtendedField<I>, ExpressionAlgebra<Double, I>> XYColumnarData<Double, Double, Double>.fitWith(optimizer: Optimizer<Double, XYFit>, processor: AutoDiffProcessor<Double, I, A>, startingPoint: Map<Symbol, Double>, vararg features: OptimizationFeature = emptyArray(), xSymbol: Symbol = Symbol.x, pointToCurveDistance: PointToCurveDistance = PointToCurveDistance.byY, pointWeight: PointWeight = PointWeight.byYSigma, model: A.(I) -> I): XYFit

Fit given dta with

FunctionOptimization

fun <T> FunctionOptimization(expression: DifferentiableExpression<T>, builder: FunctionOptimizationBuilder<T>.() -> Unit): FunctionOptimization<T>

inline fun <F : OptimizationFeature> OptimizationProblem<*>.getFeature(): F?

maximumLogLikelihood

@UnstableKMathAPI

suspend fun Optimizer<Double, FunctionOptimization<Double>>.maximumLogLikelihood(problem: XYFit): XYFit

Optimize given XY (least squares) problem using this function Optimizer. The problem is treated as maximum likelihood problem and is done via maximizing logarithmic likelihood, respecting possible weight dependency on the model and parameters.

@UnstableKMathAPI

suspend fun Optimizer<Double, FunctionOptimization<Double>>.maximumLogLikelihood(data: XYColumnarData<Double, Double, Double>, model: DifferentiableExpression<Double>, builder: XYOptimizationBuilder.() -> Unit): XYFit

suspend fun <T> DifferentiableExpression<T>.optimizeWith(optimizer: Optimizer<T, FunctionOptimization<T>>, vararg startingPoint: Pair<Symbol, T>, builder: FunctionOptimizationBuilder<T>.() -> Unit = {}): FunctionOptimization<T>

suspend fun <T> DifferentiableExpression<T>.optimizeWith(optimizer: Optimizer<T, FunctionOptimization<T>>, startingPoint: Map<Symbol, T>, builder: FunctionOptimizationBuilder<T>.() -> Unit = {}): FunctionOptimization<T>

suspend fun <T : Any> DifferentiableExpression<T>.optimizeWith(optimizer: Optimizer<T, FunctionOptimization<T>>, startingPoint: Map<Symbol, T>, vararg features: OptimizationFeature): FunctionOptimization<T>

Optimizes differentiable expression using specific optimizer form given startingPoint.

fun <T> OptimizationBuilder<T, *>.startAt(startingPoint: Map<Symbol, T>)

fun XYFit.withFeature(vararg features: OptimizationFeature): XYFit

fun <T> FunctionOptimization<T>.withFeatures(vararg newFeature: OptimizationFeature): FunctionOptimization<T>

fun XYOptimization(data: XYColumnarData<Double, Double, Double>, model: DifferentiableExpression<Double>, builder: XYOptimizationBuilder.() -> Unit): XYFit

Properties

chiSquaredOrNull

val XYFit.chiSquaredOrNull: Double?

Compute chi squared value for completed fit. Return null for incomplete fit

val <T> OptimizationProblem<T>.resultPoint: Map<Symbol, T>

resultPointOrNull

val <T> OptimizationProblem<T>.resultPointOrNull: Map<Symbol, T>?

val <T> FunctionOptimization<T>.resultValue: T

resultValueOrNull

val <T> FunctionOptimization<T>.resultValueOrNull: T?

val <T> OptimizationProblem<T>.startPoint: Map<Symbol, T>

Get the starting point for optimization. Throws error if not defined.