teldufalsari/kmath
2
0
Fork 0
forked from kscience/kmath
Code Pull Requests 1 Activity

Merge branch 'dev' into mp-samplers

Browse Source 
# Conflicts:
#	examples/src/main/kotlin/kscience/kmath/stat/DistributionDemo.kt
#	kmath-stat/src/jvmMain/kotlin/kscience/kmath/stat/distributions.kt
This commit is contained in:
Iaroslav Postovalov 2020-12-12 17:11:09 +07:00
commit 5c54640e60
No known key found for this signature in database
GPG Key ID: 46E15E4A31B3BCD7
52 changed files with 758 additions and 377 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

106
.github/workflows/gradle.yml vendored
View File

@ -1,17 +1,101 @@
name: Gradle build
on: [push]
on: [ push ]
jobs:
build:
runs-on: ubuntu-latest
build-ubuntu:
runs-on: ubuntu-20.04
steps:
- uses: actions/checkout@v1
- name: Set up JDK 11
uses: actions/setup-java@v1
with:
java-version: 11
- name: Build with Gradle
run: ./gradlew build
- uses: actions/checkout@v2
- name: Set up JDK 11
uses: actions/setup-java@v1
with:
java-version: 11
- name: Install Chrome
run: |
sudo apt install -y libappindicator1 fonts-liberation
wget https://dl.google.com/linux/direct/google-chrome-stable_current_amd64.deb
sudo dpkg -i google-chrome*.deb
- name: Cache gradle
uses: actions/cache@v2
with:
path: |
.gradle
build
~/.gradle
key: gradle
restore-keys: gradle
- name: Cache konan
uses: actions/cache@v2
with:
path: |
~/.konan/dependencies
~/.konan/kotlin-native-prebuilt-linux-*
key: ${{ runner.os }}-konan
restore-keys: ${{ runner.os }}-konan
- name: Build with Gradle
run: ./gradlew -Dorg.gradle.daemon=false --build-cache build
build-osx:
runs-on: macos-latest
steps:
- uses: actions/checkout@v2
- name: Set up JDK 11
uses: actions/setup-java@v1
with:
java-version: 11
- name: Cache gradle
uses: actions/cache@v2
with:
path: |
.gradle
build
~/.gradle
key: gradle
restore-keys: gradle
- name: Cache konan
uses: actions/cache@v2
with:
path: |
~/.konan/dependencies
~/.konan/kotlin-native-prebuilt-macos-*
key: ${{ runner.os }}-konan
restore-keys: ${{ runner.os }}-konan
- name: Build with Gradle
run: sudo ./gradlew -Dorg.gradle.daemon=false --build-cache build
build-windows:
runs-on: windows-latest
steps:
- uses: actions/checkout@v2
- name: Set up JDK 11
uses: actions/setup-java@v1
with:
java-version: 11
- name: Add msys to path
run: SETX PATH "%PATH%;C:\msys64\mingw64\bin"
- name: Cache gradle
uses: actions/cache@v2
with:
path: |
.gradle
build
~/.gradle
key: ${{ runner.os }}-gradle
restore-keys: ${{ runner.os }}-gradle
- name: Cache konan
uses: actions/cache@v2
with:
path: |
~/.konan/dependencies
~/.konan/kotlin-native-prebuilt-mingw-*
key: ${{ runner.os }}-konan
restore-keys: ${{ runner.os }}-konan
- name: Build with Gradle
run: ./gradlew --build-cache build

10
CHANGELOG.md
View File

@ -16,17 +16,22 @@
- ND4J support module submitting `NDStructure` and `NDAlgebra` over `INDArray`.
- Coroutine-deterministic Monte-Carlo scope with a random number generator.
- Some minor utilities to `kmath-for-real`.
- Generic operation result parameter to `MatrixContext`
### Changed
- Package changed from `scientifik` to `kscience.kmath`.
- Gradle version: 6.6 -> 6.7
- Gradle version: 6.6 -> 6.7.1
- Minor exceptions refactor (throwing `IllegalArgumentException` by argument checks instead of `IllegalStateException`)
- `Polynomial` secondary constructor made function.
- Kotlin version: 1.3.72 -> 1.4.20-M1
- Kotlin version: 1.3.72 -> 1.4.20
- `kmath-ast` doesn't depend on heavy `kotlin-reflect` library.
- Full autodiff refactoring based on `Symbol`
- `kmath-prob` renamed to `kmath-stat`
- Grid generators moved to `kmath-for-real`
- Use `Point<Double>` instead of specialized type in `kmath-for-real`
- Optimized dot product for buffer matrices moved to `kmath-for-real`
- EjmlMatrix context is an object
- Matrix LUP `inverse` renamed to `inverseWithLUP`
### Deprecated
@ -34,6 +39,7 @@
- `kmath-koma` module because it doesn't support Kotlin 1.4.
- Support of `legacy` JS backend (we will support only IR)
- `toGrid` method.
- Public visibility of `BufferAccessor2D`
### Fixed
- `symbol` method in `MstExtendedField` (https://github.com/mipt-npm/kmath/pull/140)

38
README.md
View File

@ -132,9 +132,17 @@ submit a feature request if you want something to be implemented first.
<hr/>
* ### [kmath-for-real](kmath-for-real)
>
> Extension module that should be used to achieve numpy-like behavior.
All operations are specialized to work with `Double` numbers without declaring algebraic contexts.
One can still use generic algebras though.
>
> **Maturity**: EXPERIMENTAL
>
> **Features:**
> - [RealVector](kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/RealVector.kt) : Numpy-like operations for Buffers/Points
> - [RealMatrix](kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/RealMatrix.kt) : Numpy-like operations for 2d real structures
> - [grids](kmath-for-real/src/commonMain/kotlin/kscience/kmath/structures/grids.kt) : Uniform grid generators
<hr/>
* ### [kmath-functions](kmath-functions)
@ -155,6 +163,12 @@ submit a feature request if you want something to be implemented first.
> **Maturity**: EXPERIMENTAL
<hr/>
* ### [kmath-kotlingrad](kmath-kotlingrad)
>
>
> **Maturity**: EXPERIMENTAL
<hr/>
* ### [kmath-memory](kmath-memory)
>
>
@ -167,7 +181,7 @@ submit a feature request if you want something to be implemented first.
> **Maturity**: EXPERIMENTAL
>
> **Features:**
> - [nd4jarraystrucure](kmath-nd4j/src/commonMain/kotlin/kscience/kmath/operations/Algebra.kt) : NDStructure wrapper for INDArray
> - [nd4jarraystructure](kmath-nd4j/src/commonMain/kotlin/kscience/kmath/operations/Algebra.kt) : NDStructure wrapper for INDArray
> - [nd4jarrayrings](kmath-nd4j/src/commonMain/kotlin/kscience/kmath/structures/NDStructure.kt) : Rings over Nd4jArrayStructure of Int and Long
> - [nd4jarrayfields](kmath-nd4j/src/commonMain/kotlin/kscience/kmath/structures/Buffers.kt) : Fields over Nd4jArrayStructure of Float and Double
@ -211,20 +225,12 @@ Release artifacts are accessible from bintray with following configuration (see
```kotlin
repositories {
jcenter()
maven("https://clojars.org/repo")
maven("https://dl.bintray.com/egor-bogomolov/astminer/")
maven("https://dl.bintray.com/hotkeytlt/maven")
maven("https://dl.bintray.com/kotlin/kotlin-eap")
maven("https://dl.bintray.com/kotlin/kotlinx")
maven("https://dl.bintray.com/mipt-npm/kscience")
maven("https://jitpack.io")
mavenCentral()
}
dependencies {
api("kscience.kmath:kmath-core:0.2.0-dev-3")
// api("kscience.kmath:kmath-core-jvm:0.2.0-dev-3") for jvm-specific version
api("kscience.kmath:kmath-core:0.2.0-dev-4")
// api("kscience.kmath:kmath-core-jvm:0.2.0-dev-4") for jvm-specific version
}
```
@ -236,15 +242,7 @@ Development builds are uploaded to the separate repository:
```kotlin
repositories {
jcenter()
maven("https://clojars.org/repo")
maven("https://dl.bintray.com/egor-bogomolov/astminer/")
maven("https://dl.bintray.com/hotkeytlt/maven")
maven("https://dl.bintray.com/kotlin/kotlin-eap")
maven("https://dl.bintray.com/kotlin/kotlinx")
maven("https://dl.bintray.com/mipt-npm/dev")
maven("https://jitpack.io")
mavenCentral()
}
```

11
examples/build.gradle.kts
View File

@ -35,6 +35,9 @@ dependencies {
implementation(project(":kmath-dimensions"))
implementation(project(":kmath-ejml"))
implementation(project(":kmath-nd4j"))
implementation(project(":kmath-for-real"))
implementation("org.deeplearning4j:deeplearning4j-core:1.0.0-beta7")
implementation("org.nd4j:nd4j-native:1.0.0-beta7")
@ -51,7 +54,11 @@ dependencies {
implementation("org.jetbrains.kotlinx:kotlinx-io:0.2.0-npm-dev-11")
implementation("org.jetbrains.kotlinx:kotlinx.benchmark.runtime:0.2.0-dev-20")
implementation("org.slf4j:slf4j-simple:1.7.30")
"benchmarksImplementation"("org.jetbrains.kotlinx:kotlinx.benchmark.runtime-jvm:0.2.0-dev-8")
// plotting
implementation("kscience.plotlykt:plotlykt-server:0.3.1-dev")
"benchmarksImplementation"("org.jetbrains.kotlinx:kotlinx.benchmark.runtime-jvm:0.2.0-dev-20")
"benchmarksImplementation"(sourceSets.main.get().output + sourceSets.main.get().runtimeClasspath)
}
@ -62,7 +69,7 @@ benchmark {
// This one matches sourceSet name above
configurations.register("fast") {
warmups = 5 // number of warmup iterations
warmups = 1 // number of warmup iterations
iterations = 3 // number of iterations
iterationTime = 500 // time in seconds per iteration
iterationTimeUnit = "ms" // time unity for iterationTime, default is seconds

2
examples/src/benchmarks/kotlin/kscience/kmath/structures/ArrayBenchmark.kt → examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/ArrayBenchmark.kt
View File

@ -1,4 +1,4 @@
package kscience.kmath.structures
package kscience.kmath.benchmarks
import org.openjdk.jmh.annotations.Benchmark
import org.openjdk.jmh.annotations.Scope

4
examples/src/benchmarks/kotlin/kscience/kmath/structures/BufferBenchmark.kt → examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/BufferBenchmark.kt
View File

@ -1,7 +1,9 @@
package kscience.kmath.structures
package kscience.kmath.benchmarks
import kscience.kmath.operations.Complex
import kscience.kmath.operations.complex
import kscience.kmath.structures.MutableBuffer
import kscience.kmath.structures.RealBuffer
import org.openjdk.jmh.annotations.Benchmark
import org.openjdk.jmh.annotations.Scope
import org.openjdk.jmh.annotations.State

50
examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/LinearAlgebraBenchmark.kt Normal file
View File

@ -0,0 +1,50 @@
package kscience.kmath.linear
import kotlinx.benchmark.Benchmark
import kscience.kmath.commons.linear.CMMatrixContext
import kscience.kmath.commons.linear.CMMatrixContext.dot
import kscience.kmath.commons.linear.inverse
import kscience.kmath.commons.linear.toCM
import kscience.kmath.ejml.EjmlMatrixContext
import kscience.kmath.ejml.inverse
import kscience.kmath.ejml.toEjml
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import org.openjdk.jmh.annotations.Scope
import org.openjdk.jmh.annotations.State
import kotlin.random.Random
@State(Scope.Benchmark)
class LinearAlgebraBenchmark {
companion object {
val random = Random(1224)
val dim = 100
//creating invertible matrix
val u = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
val l = Matrix.real(dim, dim) { i, j -> if (i >= j) random.nextDouble() else 0.0 }
val matrix = l dot u
}
@Benchmark
fun kmathLUPInversion() {
MatrixContext.real.inverseWithLUP(matrix)
}
@Benchmark
fun cmLUPInversion() {
CMMatrixContext {
val cm = matrix.toCM() //avoid overhead on conversion
inverse(cm)
}
}
@Benchmark
fun ejmlInverse() {
EjmlMatrixContext {
val km = matrix.toEjml() //avoid overhead on conversion
inverse(km)
}
}
}

60
examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/MultiplicationBenchmark.kt Normal file
View File

@ -0,0 +1,60 @@
package kscience.kmath.benchmarks
import kotlinx.benchmark.Benchmark
import kscience.kmath.commons.linear.CMMatrixContext
import kscience.kmath.commons.linear.CMMatrixContext.dot
import kscience.kmath.commons.linear.toCM
import kscience.kmath.ejml.EjmlMatrixContext
import kscience.kmath.ejml.toEjml
import kscience.kmath.linear.real
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import org.openjdk.jmh.annotations.Scope
import org.openjdk.jmh.annotations.State
import kotlin.random.Random
@State(Scope.Benchmark)
class MultiplicationBenchmark {
companion object {
val random = Random(12224)
val dim = 1000
//creating invertible matrix
val matrix1 = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
val matrix2 = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
val cmMatrix1 = matrix1.toCM()
val cmMatrix2 = matrix2.toCM()
val ejmlMatrix1 = matrix1.toEjml()
val ejmlMatrix2 = matrix2.toEjml()
}
@Benchmark
fun commonsMathMultiplication() {
CMMatrixContext.invoke {
cmMatrix1 dot cmMatrix2
}
}
@Benchmark
fun ejmlMultiplication() {
EjmlMatrixContext.invoke {
ejmlMatrix1 dot ejmlMatrix2
}
}
@Benchmark
fun ejmlMultiplicationwithConversion() {
val ejmlMatrix1 = matrix1.toEjml()
val ejmlMatrix2 = matrix2.toEjml()
EjmlMatrixContext.invoke {
ejmlMatrix1 dot ejmlMatrix2
}
}
@Benchmark
fun bufferedMultiplication() {
matrix1 dot matrix2
}
}

3
examples/src/benchmarks/kotlin/kscience/kmath/structures/NDFieldBenchmark.kt → examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/NDFieldBenchmark.kt
View File

@ -1,7 +1,8 @@
package kscience.kmath.structures
package kscience.kmath.benchmarks
import kscience.kmath.operations.RealField
import kscience.kmath.operations.invoke
import kscience.kmath.structures.*
import org.openjdk.jmh.annotations.Benchmark
import org.openjdk.jmh.annotations.Scope
import org.openjdk.jmh.annotations.State

5
examples/src/benchmarks/kotlin/kscience/kmath/structures/ViktorBenchmark.kt → examples/src/benchmarks/kotlin/kscience/kmath/benchmarks/ViktorBenchmark.kt
View File

@ -1,7 +1,10 @@
package kscience.kmath.structures
package kscience.kmath.benchmarks
import kscience.kmath.operations.RealField
import kscience.kmath.operations.invoke
import kscience.kmath.structures.BufferedNDField
import kscience.kmath.structures.NDField
import kscience.kmath.structures.RealNDField
import kscience.kmath.viktor.ViktorNDField
import org.jetbrains.bio.viktor.F64Array
import org.openjdk.jmh.annotations.Benchmark

11
examples/src/benchmarks/kotlin/kscience/kmath/utils/utils.kt
View File

@ -1,11 +0,0 @@
package kscience.kmath.utils
import kotlin.contracts.InvocationKind
import kotlin.contracts.contract
import kotlin.system.measureTimeMillis
internal inline fun measureAndPrint(title: String, block: () -> Unit) {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
val time = measureTimeMillis(block)
println("$title completed in $time millis")
}

102
examples/src/main/kotlin/kscience/kmath/commons/fit/fitWithAutoDiff.kt Normal file
View File

@ -0,0 +1,102 @@
package kscience.kmath.commons.fit
import kotlinx.html.br
import kotlinx.html.h3
import kscience.kmath.commons.optimization.chiSquared
import kscience.kmath.commons.optimization.minimize
import kscience.kmath.expressions.symbol
import kscience.kmath.real.RealVector
import kscience.kmath.real.map
import kscience.kmath.real.step
import kscience.kmath.stat.*
import kscience.kmath.structures.asIterable
import kscience.kmath.structures.toList
import kscience.plotly.*
import kscience.plotly.models.ScatterMode
import kscience.plotly.models.TraceValues
import kotlin.math.pow
import kotlin.math.sqrt
//Forward declaration of symbols that will be used in expressions.
// This declaration is required for
private val a by symbol
private val b by symbol
private val c by symbol
/**
* Shortcut to use buffers in plotly
*/
operator fun TraceValues.invoke(vector: RealVector) {
numbers = vector.asIterable()
}
/**
* Least squares fie with auto-differentiation. Uses `kmath-commons` and `kmath-for-real` modules.
*/
fun main() {
//A generator for a normally distributed values
val generator = Distribution.normal()
//A chain/flow of random values with the given seed
val chain = generator.sample(RandomGenerator.default(112667))
//Create a uniformly distributed x values like numpy.arrange
val x = 1.0..100.0 step 1.0
//Perform an operation on each x value (much more effective, than numpy)
val y = x.map {
val value = it.pow(2) + it + 1
value + chain.nextDouble() * sqrt(value)
}
// this will also work, but less effective:
// val y = x.pow(2)+ x + 1 + chain.nextDouble()
// create same errors for all xs
val yErr = y.map { sqrt(it) }//RealVector.same(x.size, sigma)
// compute differentiable chi^2 sum for given model ax^2 + bx + c
val chi2 = Fitting.chiSquared(x, y, yErr) { x1 ->
//bind variables to autodiff context
val a = bind(a)
val b = bind(b)
//Include default value for c if it is not provided as a parameter
val c = bindOrNull(c) ?: one
a * x1.pow(2) + b * x1 + c
}
//minimize the chi^2 in given starting point. Derivatives are not required, they are already included.
val result: OptimizationResult<Double> = chi2.minimize(a to 1.5, b to 0.9, c to 1.0)
//display a page with plot and numerical results
val page = Plotly.page {
plot {
scatter {
mode = ScatterMode.markers
x(x)
y(y)
error_y {
array = yErr.toList()
}
name = "data"
}
scatter {
mode = ScatterMode.lines
x(x)
y(x.map { result.point[a]!! * it.pow(2) + result.point[b]!! * it + 1 })
name = "fit"
}
}
br()
h3{
+"Fit result: $result"
}
h3{
+"Chi2/dof = ${result.value / (x.size - 3)}"
}
}
page.makeFile()
}

50
examples/src/main/kotlin/kscience/kmath/linear/LinearAlgebraBenchmark.kt
View File

@ -1,50 +0,0 @@
package kscience.kmath.linear
import kscience.kmath.commons.linear.CMMatrixContext
import kscience.kmath.commons.linear.inverse
import kscience.kmath.commons.linear.toCM
import kscience.kmath.ejml.EjmlMatrixContext
import kscience.kmath.ejml.inverse
import kscience.kmath.operations.RealField
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import kotlin.random.Random
import kotlin.system.measureTimeMillis
fun main() {
val random = Random(1224)
val dim = 100
//creating invertible matrix
val u = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
val l = Matrix.real(dim, dim) { i, j -> if (i >= j) random.nextDouble() else 0.0 }
val matrix = l dot u
val n = 5000 // iterations
MatrixContext.real {
repeat(50) { inverse(matrix) }
val inverseTime = measureTimeMillis { repeat(n) { inverse(matrix) } }
println("[kmath] Inversion of $n matrices $dim x $dim finished in $inverseTime millis")
}
//commons-math
val commonsTime = measureTimeMillis {
CMMatrixContext {
val cm = matrix.toCM() //avoid overhead on conversion
repeat(n) { inverse(cm) }
}
}
println("[commons-math] Inversion of $n matrices $dim x $dim finished in $commonsTime millis")
val ejmlTime = measureTimeMillis {
(EjmlMatrixContext(RealField)) {
val km = matrix.toEjml() //avoid overhead on conversion
repeat(n) { inverse(km) }
}
}
println("[ejml] Inversion of $n matrices $dim x $dim finished in $ejmlTime millis")
}

38
examples/src/main/kotlin/kscience/kmath/linear/MultiplicationBenchmark.kt
View File

@ -1,38 +0,0 @@
package kscience.kmath.linear
import kscience.kmath.commons.linear.CMMatrixContext
import kscience.kmath.commons.linear.toCM
import kscience.kmath.ejml.EjmlMatrixContext
import kscience.kmath.operations.RealField
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import kotlin.random.Random
import kotlin.system.measureTimeMillis
fun main() {
val random = Random(12224)
val dim = 1000
//creating invertible matrix
val matrix1 = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
val matrix2 = Matrix.real(dim, dim) { i, j -> if (i <= j) random.nextDouble() else 0.0 }
// //warmup
// matrix1 dot matrix2
CMMatrixContext {
val cmMatrix1 = matrix1.toCM()
val cmMatrix2 = matrix2.toCM()
val cmTime = measureTimeMillis { cmMatrix1 dot cmMatrix2 }
println("CM implementation time: $cmTime")
}
(EjmlMatrixContext(RealField)) {
val ejmlMatrix1 = matrix1.toEjml()
val ejmlMatrix2 = matrix2.toEjml()
val ejmlTime = measureTimeMillis { ejmlMatrix1 dot ejmlMatrix2 }
println("EJML implementation time: $ejmlTime")
}
val genericTime = measureTimeMillis { val res = matrix1 dot matrix2 }
println("Generic implementation time: $genericTime")
}

10
examples/src/main/kotlin/kscience/kmath/stat/DistributionDemo.kt
View File

@ -3,11 +3,15 @@ package kscience.kmath.stat
import kotlinx.coroutines.runBlocking
import kscience.kmath.chains.Chain
import kscience.kmath.chains.collectWithState
import kscience.kmath.stat.RandomGenerator
import kscience.kmath.stat.samplers.ZigguratNormalizedGaussianSampler
/**
* The state of distribution averager.
*/
private data class AveragingChainState(var num: Int = 0, var value: Double = 0.0)
/**
* Averaging.
*/
private fun Chain<Double>.mean(): Chain<Double> = collectWithState(AveragingChainState(), { it.copy() }) { chain ->
val next = chain.next()
num++
@ -17,7 +21,7 @@ private fun Chain<Double>.mean(): Chain<Double> = collectWithState(AveragingChai
fun main() {
val normal = ZigguratNormalizedGaussianSampler.of()
val normal = NormalDistribution()
val chain = normal.sample(RandomGenerator.default).mean()
runBlocking {

4
kmath-ast/build.gradle.kts
View File

@ -17,3 +17,7 @@ kotlin.sourceSets {
}
}
}
readme{
maturity = ru.mipt.npm.gradle.Maturity.PROTOTYPE
}

4
kmath-ast/src/jvmMain/kotlin/kscience/kmath/ast/parser.kt
View File

@ -83,11 +83,10 @@ public object ArithmeticsEvaluator : Grammar<MST>() {
}
/**
* Tries to parse the string into [MST].
* Tries to parse the string into [MST]. Returns [ParseResult] representing expression or error.
*
* @receiver the string to parse.
* @return the [MST] node.
* @author Alexander Nozik
*/
public fun String.tryParseMath(): ParseResult<MST> = ArithmeticsEvaluator.tryParseToEnd(this)
@ -96,6 +95,5 @@ public fun String.tryParseMath(): ParseResult<MST> = ArithmeticsEvaluator.tryPar
*
* @receiver the string to parse.
* @return the [MST] node.
* @author Alexander Nozik
*/
public fun String.parseMath(): MST = ArithmeticsEvaluator.parseToEnd(this)

7
kmath-commons/src/main/kotlin/kscience/kmath/commons/linear/CMMatrix.kt
View File

@ -5,8 +5,7 @@ import kscience.kmath.structures.Matrix
import kscience.kmath.structures.NDStructure
import org.apache.commons.math3.linear.*
public class CMMatrix(public val origin: RealMatrix, features: Set<MatrixFeature>? = null) :
FeaturedMatrix<Double> {
public class CMMatrix(public val origin: RealMatrix, features: Set<MatrixFeature>? = null) : FeaturedMatrix<Double> {
public override val rowNum: Int get() = origin.rowDimension
public override val colNum: Int get() = origin.columnDimension
@ -55,7 +54,7 @@ public fun Point<Double>.toCM(): CMVector = if (this is CMVector) this else {
public fun RealVector.toPoint(): CMVector = CMVector(this)
public object CMMatrixContext : MatrixContext<Double> {
public object CMMatrixContext : MatrixContext<Double, CMMatrix> {
public override fun produce(rows: Int, columns: Int, initializer: (i: Int, j: Int) -> Double): CMMatrix {
val array = Array(rows) { i -> DoubleArray(columns) { j -> initializer(i, j) } }
return CMMatrix(Array2DRowRealMatrix(array))
@ -79,7 +78,7 @@ public object CMMatrixContext : MatrixContext<Double> {
public override fun multiply(a: Matrix<Double>, k: Number): CMMatrix =
CMMatrix(a.toCM().origin.scalarMultiply(k.toDouble()))
public override operator fun Matrix<Double>.times(value: Double): Matrix<Double> =
public override operator fun Matrix<Double>.times(value: Double): CMMatrix =
produce(rowNum, colNum) { i, j -> get(i, j) * value }
}

6
kmath-core/README.md
View File

@ -12,7 +12,7 @@ The core features of KMath:
> #### Artifact:
>
> This module artifact: `kscience.kmath:kmath-core:0.2.0-dev-3`.
> This module artifact: `kscience.kmath:kmath-core:0.2.0-dev-4`.
>
> Bintray release version: [ ![Download](https://api.bintray.com/packages/mipt-npm/kscience/kmath-core/images/download.svg) ](https://bintray.com/mipt-npm/kscience/kmath-core/_latestVersion)
>
@ -30,7 +30,7 @@ The core features of KMath:
> }
>
> dependencies {
> implementation 'kscience.kmath:kmath-core:0.2.0-dev-3'
> implementation 'kscience.kmath:kmath-core:0.2.0-dev-4'
> }
> ```
> **Gradle Kotlin DSL:**
@ -44,6 +44,6 @@ The core features of KMath:
> }
>
> dependencies {
> implementation("kscience.kmath:kmath-core:0.2.0-dev-3")
> implementation("kscience.kmath:kmath-core:0.2.0-dev-4")
> }
> ```

4
kmath-core/build.gradle.kts
View File

@ -1,5 +1,3 @@
import ru.mipt.npm.gradle.Maturity
plugins {
id("ru.mipt.npm.mpp")
id("ru.mipt.npm.native")
@ -13,7 +11,7 @@ kotlin.sourceSets.commonMain {
readme {
description = "Core classes, algebra definitions, basic linear algebra"
maturity = Maturity.DEVELOPMENT
maturity = ru.mipt.npm.gradle.Maturity.DEVELOPMENT
propertyByTemplate("artifact", rootProject.file("docs/templates/ARTIFACT-TEMPLATE.md"))
feature(

41
kmath-core/src/commonMain/kotlin/kscience/kmath/linear/BufferMatrix.kt
View File

@ -9,8 +9,8 @@ import kscience.kmath.structures.*
*/
public class BufferMatrixContext<T : Any, R : Ring<T>>(
public override val elementContext: R,
private val bufferFactory: BufferFactory<T>
) : GenericMatrixContext<T, R> {
private val bufferFactory: BufferFactory<T>,
) : GenericMatrixContext<T, R, BufferMatrix<T>> {
public override fun produce(rows: Int, columns: Int, initializer: (i: Int, j: Int) -> T): BufferMatrix<T> {
val buffer = bufferFactory(rows * columns) { offset -> initializer(offset / columns, offset % columns) }
return BufferMatrix(rows, columns, buffer)
@ -22,15 +22,15 @@ public class BufferMatrixContext<T : Any, R : Ring<T>>(
}
@Suppress("OVERRIDE_BY_INLINE")
public object RealMatrixContext : GenericMatrixContext<Double, RealField> {
public object RealMatrixContext : GenericMatrixContext<Double, RealField, BufferMatrix<Double>> {
public override val elementContext: RealField
get() = RealField
public override inline fun produce(
rows: Int,
columns: Int,
initializer: (i: Int, j: Int) -> Double
): Matrix<Double> {
initializer: (i: Int, j: Int) -> Double,
): BufferMatrix<Double> {
val buffer = RealBuffer(rows * columns) { offset -> initializer(offset / columns, offset % columns) }
return BufferMatrix(rows, columns, buffer)
}
@ -43,15 +43,15 @@ public class BufferMatrix<T : Any>(
public override val rowNum: Int,
public override val colNum: Int,
public val buffer: Buffer<out T>,
public override val features: Set<MatrixFeature> = emptySet()
public override val features: Set<MatrixFeature> = emptySet(),
) : FeaturedMatrix<T> {
override val shape: IntArray
get() = intArrayOf(rowNum, colNum)
init {
require(buffer.size == rowNum * colNum) { "Dimension mismatch for matrix structure" }
}
override val shape: IntArray get() = intArrayOf(rowNum, colNum)
public override fun suggestFeature(vararg features: MatrixFeature): BufferMatrix<T> =
BufferMatrix(rowNum, colNum, buffer, this.features + features)
@ -86,28 +86,3 @@ public class BufferMatrix<T : Any>(
else "Matrix(rowsNum = $rowNum, colNum = $colNum, features=$features)"
}
}
/**
* Optimized dot product for real matrices
*/
public infix fun BufferMatrix<Double>.dot(other: BufferMatrix<Double>): BufferMatrix<Double> {
require(colNum == other.rowNum) { "Matrix dot operation dimension mismatch: ($rowNum, $colNum) x (${other.rowNum}, ${other.colNum})" }
val array = DoubleArray(this.rowNum * other.colNum)
//convert to array to insure there is not memory indirection
fun Buffer<out Double>.unsafeArray() = if (this is RealBuffer)
array
else
DoubleArray(size) { get(it) }
val a = this.buffer.unsafeArray()
val b = other.buffer.unsafeArray()
for (i in (0 until rowNum))
for (j in (0 until other.colNum))
for (k in (0 until colNum))
array[i * other.colNum + j] += a[i * colNum + k] * b[k * other.colNum + j]
val buffer = RealBuffer(array)
return BufferMatrix(rowNum, other.colNum, buffer)
}

13
kmath-core/src/commonMain/kotlin/kscience/kmath/linear/FeaturedMatrix.kt
View File

@ -27,9 +27,8 @@ public interface FeaturedMatrix<T : Any> : Matrix<T> {
public inline fun Structure2D.Companion.real(
rows: Int,
columns: Int,
initializer: (Int, Int) -> Double
): Matrix<Double> =
MatrixContext.real.produce(rows, columns, initializer)
initializer: (Int, Int) -> Double,
): BufferMatrix<Double> = MatrixContext.real.produce(rows, columns, initializer)
/**
* Build a square matrix from given elements.
@ -58,7 +57,7 @@ public inline fun <reified T : Any> Matrix<*>.getFeature(): T? =
/**
* Diagonal matrix of ones. The matrix is virtual no actual matrix is created
*/
public fun <T : Any, R : Ring<T>> GenericMatrixContext<T, R>.one(rows: Int, columns: Int): FeaturedMatrix<T> =
public fun <T : Any, R : Ring<T>> GenericMatrixContext<T, R, *>.one(rows: Int, columns: Int): FeaturedMatrix<T> =
VirtualMatrix(rows, columns, DiagonalFeature) { i, j ->
if (i == j) elementContext.one else elementContext.zero
}
@ -67,7 +66,7 @@ public fun <T : Any, R : Ring<T>> GenericMatrixContext<T, R>.one(rows: Int, colu
/**
* A virtual matrix of zeroes
*/
public fun <T : Any, R : Ring<T>> GenericMatrixContext<T, R>.zero(rows: Int, columns: Int): FeaturedMatrix<T> =
public fun <T : Any, R : Ring<T>> GenericMatrixContext<T, R, *>.zero(rows: Int, columns: Int): FeaturedMatrix<T> =
VirtualMatrix(rows, columns) { _, _ -> elementContext.zero }
public class TransposedFeature<T : Any>(public val original: Matrix<T>) : MatrixFeature
@ -81,6 +80,4 @@ public fun <T : Any> Matrix<T>.transpose(): Matrix<T> {
rowNum,
setOf(TransposedFeature(this))
) { i, j -> get(j, i) }
}
public infix fun Matrix<Double>.dot(other: Matrix<Double>): Matrix<Double> = with(MatrixContext.real) { dot(other) }
}

86
kmath-core/src/commonMain/kotlin/kscience/kmath/linear/LUPDecomposition.kt
View File

@ -1,25 +1,18 @@
package kscience.kmath.linear
import kscience.kmath.operations.Field
import kscience.kmath.operations.RealField
import kscience.kmath.operations.Ring
import kscience.kmath.operations.invoke
import kscience.kmath.structures.BufferAccessor2D
import kscience.kmath.structures.Matrix
import kscience.kmath.structures.Structure2D
import kotlin.reflect.KClass
import kscience.kmath.operations.*
import kscience.kmath.structures.*
/**
* Common implementation of [LUPDecompositionFeature]
*/
public class LUPDecomposition<T : Any>(
public val context: GenericMatrixContext<T, out Field<T>>,
public val context: MatrixContext<T, FeaturedMatrix<T>>,
public val elementContext: Field<T>,
public val lu: Structure2D<T>,
public val pivot: IntArray,
private val even: Boolean
private val even: Boolean,
) : LUPDecompositionFeature<T>, DeterminantFeature<T> {
public val elementContext: Field<T>
get() = context.elementContext
/**
* Returns the matrix L of the decomposition.
@ -64,23 +57,25 @@ public class LUPDecomposition<T : Any>(
}
public fun <T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.abs(value: T): T =
@PublishedApi
internal fun <T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F, *>.abs(value: T): T =
if (value > elementContext.zero) value else elementContext { -value }
/**
* Create a lup decomposition of generic matrix
* Create a lup decomposition of generic matrix.
*/
public inline fun <T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.lup(
type: KClass<T>,
public fun <T : Comparable<T>> MatrixContext<T, FeaturedMatrix<T>>.lup(
factory: MutableBufferFactory<T>,
elementContext: Field<T>,
matrix: Matrix<T>,
checkSingular: (T) -> Boolean
checkSingular: (T) -> Boolean,
): LUPDecomposition<T> {
require(matrix.rowNum == matrix.colNum) { "LU decomposition supports only square matrices" }
val m = matrix.colNum
val pivot = IntArray(matrix.rowNum)
//TODO just waits for KEEP-176
BufferAccessor2D(type, matrix.rowNum, matrix.colNum).run {
BufferAccessor2D(matrix.rowNum, matrix.colNum, factory).run {
elementContext {
val lu = create(matrix)
@ -112,14 +107,14 @@ public inline fun <T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.l
luRow[col] = sum
// maintain best permutation choice
if (this@lup.abs(sum) > largest) {
largest = this@lup.abs(sum)
if (abs(sum) > largest) {
largest = abs(sum)
max = row
}
}
// Singularity check
check(!checkSingular(this@lup.abs(lu[max, col]))) { "The matrix is singular" }
check(!checkSingular(abs(lu[max, col]))) { "The matrix is singular" }
// Pivot if necessary
if (max != col) {
@ -143,23 +138,23 @@ public inline fun <T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.l
for (row in col + 1 until m) lu[row, col] /= luDiag
}
return LUPDecomposition(this@lup, lu.collect(), pivot, even)
return LUPDecomposition(this@lup, elementContext, lu.collect(), pivot, even)
}
}
}
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.lup(
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F, FeaturedMatrix<T>>.lup(
matrix: Matrix<T>,
checkSingular: (T) -> Boolean
): LUPDecomposition<T> = lup(T::class, matrix, checkSingular)
noinline checkSingular: (T) -> Boolean,
): LUPDecomposition<T> = lup(MutableBuffer.Companion::auto, elementContext, matrix, checkSingular)
public fun GenericMatrixContext<Double, RealField>.lup(matrix: Matrix<Double>): LUPDecomposition<Double> =
lup(Double::class, matrix) { it < 1e-11 }
public fun MatrixContext<Double, FeaturedMatrix<Double>>.lup(matrix: Matrix<Double>): LUPDecomposition<Double> =
lup(Buffer.Companion::real, RealField, matrix) { it < 1e-11 }
public fun <T : Any> LUPDecomposition<T>.solve(type: KClass<T>, matrix: Matrix<T>): Matrix<T> {
public fun <T : Any> LUPDecomposition<T>.solveWithLUP(factory: MutableBufferFactory<T>, matrix: Matrix<T>): FeaturedMatrix<T> {
require(matrix.rowNum == pivot.size) { "Matrix dimension mismatch. Expected ${pivot.size}, but got ${matrix.colNum}" }
BufferAccessor2D(type, matrix.rowNum, matrix.colNum).run {
BufferAccessor2D(matrix.rowNum, matrix.colNum, factory).run {
elementContext {
// Apply permutations to b
val bp = create { _, _ -> zero }
@ -201,27 +196,34 @@ public fun <T : Any> LUPDecomposition<T>.solve(type: KClass<T>, matrix: Matrix<T
}
}
public inline fun <reified T : Any> LUPDecomposition<T>.solve(matrix: Matrix<T>): Matrix<T> = solve(T::class, matrix)
public inline fun <reified T : Any> LUPDecomposition<T>.solveWithLUP(matrix: Matrix<T>): Matrix<T> =
solveWithLUP(MutableBuffer.Companion::auto, matrix)
/**
* Solve a linear equation **a*x = b**
* Solve a linear equation **a*x = b** using LUP decomposition
*/
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.solve(
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F, FeaturedMatrix<T>>.solveWithLUP(
a: Matrix<T>,
b: Matrix<T>,
checkSingular: (T) -> Boolean
): Matrix<T> {
noinline bufferFactory: MutableBufferFactory<T> = MutableBuffer.Companion::auto,
noinline checkSingular: (T) -> Boolean,
): FeaturedMatrix<T> {
// Use existing decomposition if it is provided by matrix
val decomposition = a.getFeature() ?: lup(T::class, a, checkSingular)
return decomposition.solve(T::class, b)
val decomposition = a.getFeature() ?: lup(bufferFactory, elementContext, a, checkSingular)
return decomposition.solveWithLUP(bufferFactory, b)
}
public fun RealMatrixContext.solve(a: Matrix<Double>, b: Matrix<Double>): Matrix<Double> = solve(a, b) { it < 1e-11 }
public fun RealMatrixContext.solveWithLUP(a: Matrix<Double>, b: Matrix<Double>): FeaturedMatrix<Double> =
solveWithLUP(a, b) { it < 1e-11 }
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F>.inverse(
public inline fun <reified T : Comparable<T>, F : Field<T>> GenericMatrixContext<T, F, FeaturedMatrix<T>>.inverseWithLUP(
matrix: Matrix<T>,
checkSingular: (T) -> Boolean
): Matrix<T> = solve(matrix, one(matrix.rowNum, matrix.colNum), checkSingular)
noinline bufferFactory: MutableBufferFactory<T> = MutableBuffer.Companion::auto,
noinline checkSingular: (T) -> Boolean,
): FeaturedMatrix<T> = solveWithLUP(matrix, one(matrix.rowNum, matrix.colNum), bufferFactory, checkSingular)
public fun RealMatrixContext.inverse(matrix: Matrix<Double>): Matrix<Double> =
solve(matrix, one(matrix.rowNum, matrix.colNum)) { it < 1e-11 }
/**
* Inverses a square matrix using LUP decomposition. Non square matrix will throw a error.
*/
public fun RealMatrixContext.inverseWithLUP(matrix: Matrix<Double>): FeaturedMatrix<Double> =
solveWithLUP(matrix, one(matrix.rowNum, matrix.colNum), Buffer.Companion::real) { it < 1e-11 }

37
kmath-core/src/commonMain/kotlin/kscience/kmath/linear/MatrixContext.kt
View File

@ -12,15 +12,16 @@ import kscience.kmath.structures.asSequence
/**
* Basic operations on matrices. Operates on [Matrix]
*/
public interface MatrixContext<T : Any> : SpaceOperations<Matrix<T>> {
public interface MatrixContext<T : Any, out M : Matrix<T>> : SpaceOperations<Matrix<T>> {
/**
* Produce a matrix with this context and given dimensions
*/
public fun produce(rows: Int, columns: Int, initializer: (i: Int, j: Int) -> T): Matrix<T>
public fun produce(rows: Int, columns: Int, initializer: (i: Int, j: Int) -> T): M
public override fun binaryOperation(operation: String, left: Matrix<T>, right: Matrix<T>): Matrix<T> = when (operation) {
@Suppress("UNCHECKED_CAST")
public override fun binaryOperation(operation: String, left: Matrix<T>, right: Matrix<T>): M = when (operation) {
"dot" -> left dot right
else -> super.binaryOperation(operation, left, right)
else -> super.binaryOperation(operation, left, right) as M
}
/**
@ -30,7 +31,7 @@ public interface MatrixContext<T : Any> : SpaceOperations<Matrix<T>> {
* @param other the multiplier.
* @return the dot product.
*/
public infix fun Matrix<T>.dot(other: Matrix<T>): Matrix<T>
public infix fun Matrix<T>.dot(other: Matrix<T>): M
/**
* Computes the dot product of this matrix and a vector.
@ -48,7 +49,7 @@ public interface MatrixContext<T : Any> : SpaceOperations<Matrix<T>> {
* @param value the multiplier.
* @receiver the product.
*/
public operator fun Matrix<T>.times(value: T): Matrix<T>
public operator fun Matrix<T>.times(value: T): M
/**
* Multiplies an element by a matrix of it.
@ -57,7 +58,7 @@ public interface MatrixContext<T : Any> : SpaceOperations<Matrix<T>> {
* @param value the multiplier.
* @receiver the product.
*/
public operator fun T.times(m: Matrix<T>): Matrix<T> = m * this
public operator fun T.times(m: Matrix<T>): M = m * this
public companion object {
/**
@ -70,18 +71,18 @@ public interface MatrixContext<T : Any> : SpaceOperations<Matrix<T>> {
*/
public fun <T : Any, R : Ring<T>> buffered(
ring: R,
bufferFactory: BufferFactory<T> = Buffer.Companion::boxing
): GenericMatrixContext<T, R> = BufferMatrixContext(ring, bufferFactory)
bufferFactory: BufferFactory<T> = Buffer.Companion::boxing,
): GenericMatrixContext<T, R, BufferMatrix<T>> = BufferMatrixContext(ring, bufferFactory)
/**
* Automatic buffered matrix, unboxed if it is possible
*/
public inline fun <reified T : Any, R : Ring<T>> auto(ring: R): GenericMatrixContext<T, R> =
public inline fun <reified T : Any, R : Ring<T>> auto(ring: R): GenericMatrixContext<T, R, BufferMatrix<T>> =
buffered(ring, Buffer.Companion::auto)
}
}
public interface GenericMatrixContext<T : Any, R : Ring<T>> : MatrixContext<T> {
public interface GenericMatrixContext<T : Any, R : Ring<T>, out M : Matrix<T>> : MatrixContext<T, M> {
/**
* The ring context for matrix elements
*/
@ -92,7 +93,7 @@ public interface GenericMatrixContext<T : Any, R : Ring<T>> : MatrixContext<T> {
*/
public fun point(size: Int, initializer: (Int) -> T): Point<T>
public override infix fun Matrix<T>.dot(other: Matrix<T>): Matrix<T> {
public override infix fun Matrix<T>.dot(other: Matrix<T>): M {
//TODO add typed error
require(colNum == other.rowNum) { "Matrix dot operation dimension mismatch: ($rowNum, $colNum) x (${other.rowNum}, ${other.colNum})" }
@ -113,10 +114,10 @@ public interface GenericMatrixContext<T : Any, R : Ring<T>> : MatrixContext<T> {
}
}
public override operator fun Matrix<T>.unaryMinus(): Matrix<T> =
public override operator fun Matrix<T>.unaryMinus(): M =
produce(rowNum, colNum) { i, j -> elementContext { -get(i, j) } }
public override fun add(a: Matrix<T>, b: Matrix<T>): Matrix<T> {
public override fun add(a: Matrix<T>, b: Matrix<T>): M {
require(a.rowNum == b.rowNum && a.colNum == b.colNum) {
"Matrix operation dimension mismatch. [${a.rowNum},${a.colNum}] + [${b.rowNum},${b.colNum}]"
}
@ -124,7 +125,7 @@ public interface GenericMatrixContext<T : Any, R : Ring<T>> : MatrixContext<T> {
return produce(a.rowNum, a.colNum) { i, j -> elementContext { a[i, j] + b[i, j] } }
}
public override operator fun Matrix<T>.minus(b: Matrix<T>): Matrix<T> {
public override operator fun Matrix<T>.minus(b: Matrix<T>): M {
require(rowNum == b.rowNum && colNum == b.colNum) {
"Matrix operation dimension mismatch. [$rowNum,$colNum] - [${b.rowNum},${b.colNum}]"
}
@ -132,11 +133,11 @@ public interface GenericMatrixContext<T : Any, R : Ring<T>> : MatrixContext<T> {
return produce(rowNum, colNum) { i, j -> elementContext { get(i, j) + b[i, j] } }
}
public override fun multiply(a: Matrix<T>, k: Number): Matrix<T> =
public override fun multiply(a: Matrix<T>, k: Number): M =
produce(a.rowNum, a.colNum) { i, j -> elementContext { a[i, j] * k } }
public operator fun Number.times(matrix: FeaturedMatrix<T>): Matrix<T> = matrix * this
public operator fun Number.times(matrix: FeaturedMatrix<T>): M = multiply(matrix, this)
public override operator fun Matrix<T>.times(value: T): Matrix<T> =
public override operator fun Matrix<T>.times(value: T): M =
produce(rowNum, colNum) { i, j -> elementContext { get(i, j) * value } }
}

8
kmath-core/src/commonMain/kotlin/kscience/kmath/linear/VectorSpace.kt
View File

@ -15,7 +15,7 @@ public interface VectorSpace<T : Any, S : Space<T>> : Space<Point<T>> {
public val space: S
override val zero: Point<T> get() = produce { space.zero }
public fun produce(initializer: (Int) -> T): Point<T>
public fun produce(initializer: S.(Int) -> T): Point<T>
/**
* Produce a space-element of this vector space for expressions
@ -48,7 +48,7 @@ public interface VectorSpace<T : Any, S : Space<T>> : Space<Point<T>> {
public fun <T : Any, S : Space<T>> buffered(
size: Int,
space: S,
bufferFactory: BufferFactory<T> = Buffer.Companion::boxing
bufferFactory: BufferFactory<T> = Buffer.Companion::boxing,
): BufferVectorSpace<T, S> = BufferVectorSpace(size, space, bufferFactory)
/**
@ -63,8 +63,8 @@ public interface VectorSpace<T : Any, S : Space<T>> : Space<Point<T>> {
public class BufferVectorSpace<T : Any, S : Space<T>>(
override val size: Int,
override val space: S,
public val bufferFactory: BufferFactory<T>
public val bufferFactory: BufferFactory<T>,
) : VectorSpace<T, S> {
override fun produce(initializer: (Int) -> T): Buffer<T> = bufferFactory(size, initializer)
override fun produce(initializer: S.(Int) -> T): Buffer<T> = bufferFactory(size) { space.initializer(it) }
//override fun produceElement(initializer: (Int) -> T): Vector<T, S> = BufferVector(this, produce(initializer))
}

4
kmath-core/src/commonMain/kotlin/kscience/kmath/misc/annotations.kt Normal file
View File

@ -0,0 +1,4 @@
package kscience.kmath.misc
@RequiresOptIn("This API is unstable and could change in future", RequiresOptIn.Level.WARNING)
public annotation class UnstableKMathAPI

5
kmath-core/src/commonMain/kotlin/kscience/kmath/operations/AlgebraExtensions.kt
View File

@ -38,6 +38,11 @@ public fun <T> Space<T>.average(data: Iterable<T>): T = sum(data) / data.count()
*/
public fun <T> Space<T>.average(data: Sequence<T>): T = sum(data) / data.count()
/**
* Absolute of the comparable [value]
*/
public fun <T : Comparable<T>> Space<T>.abs(value: T): T = if (value > zero) value else -value
/**
* Returns the sum of all elements in the iterable in provided space.
*

22
kmath-core/src/commonMain/kotlin/kscience/kmath/structures/BufferAccessor2D.kt
View File

@ -1,25 +1,31 @@
package kscience.kmath.structures
import kotlin.reflect.KClass
/**
* A context that allows to operate on a [MutableBuffer] as on 2d array
*/
public class BufferAccessor2D<T : Any>(public val type: KClass<T>, public val rowNum: Int, public val colNum: Int) {
internal class BufferAccessor2D<T : Any>(
public val rowNum: Int,
public val colNum: Int,
val factory: MutableBufferFactory<T>,
) {
public operator fun Buffer<T>.get(i: Int, j: Int): T = get(i + colNum * j)
public operator fun MutableBuffer<T>.set(i: Int, j: Int, value: T) {
set(i + colNum * j, value)
}
public inline fun create(init: (i: Int, j: Int) -> T): MutableBuffer<T> =
MutableBuffer.auto(type, rowNum * colNum) { offset -> init(offset / colNum, offset % colNum) }
public inline fun create(crossinline init: (i: Int, j: Int) -> T): MutableBuffer<T> =
factory(rowNum * colNum) { offset -> init(offset / colNum, offset % colNum) }
public fun create(mat: Structure2D<T>): MutableBuffer<T> = create { i, j -> mat[i, j] }
//TODO optimize wrapper
public fun MutableBuffer<T>.collect(): Structure2D<T> =
NDStructure.auto(type, rowNum, colNum) { (i, j) -> get(i, j) }.as2D()
public fun MutableBuffer<T>.collect(): Structure2D<T> = NDStructure.build(
DefaultStrides(intArrayOf(rowNum, colNum)),
factory
) { (i, j) ->
get(i, j)
}.as2D()
public inner class Row(public val buffer: MutableBuffer<T>, public val rowIndex: Int) : MutableBuffer<T> {
override val size: Int get() = colNum
@ -30,7 +36,7 @@ public class BufferAccessor2D<T : Any>(public val type: KClass<T>, public val ro
buffer[rowIndex, index] = value
}
override fun copy(): MutableBuffer<T> = MutableBuffer.auto(type, colNum) { get(it) }
override fun copy(): MutableBuffer<T> = factory(colNum) { get(it) }
override operator fun iterator(): Iterator<T> = (0 until colNum).map(::get).iterator()
}

5
kmath-core/src/commonMain/kotlin/kscience/kmath/structures/Buffers.kt
View File

@ -102,6 +102,11 @@ public fun <T> Buffer<T>.asSequence(): Sequence<T> = Sequence(::iterator)
*/
public fun <T> Buffer<T>.asIterable(): Iterable<T> = Iterable(::iterator)
/**
* Converts this [Buffer] to a new [List]
*/
public fun <T> Buffer<T>.toList(): List<T> = asSequence().toList()
/**
* Returns an [IntRange] of the valid indices for this [Buffer].
*/

3
kmath-core/src/commonTest/kotlin/kscience/kmath/linear/MatrixTest.kt
View File

@ -1,5 +1,6 @@
package kscience.kmath.linear
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import kscience.kmath.structures.NDStructure
import kscience.kmath.structures.as2D
@ -38,7 +39,7 @@ class MatrixTest {
infix fun Matrix<Double>.pow(power: Int): Matrix<Double> {
var res = this
repeat(power - 1) {
res = res dot this
res = RealMatrixContext.invoke { res dot this@pow }
}
return res
}

4
kmath-core/src/commonTest/kotlin/kscience/kmath/linear/RealLUSolverTest.kt
View File

@ -9,7 +9,7 @@ class RealLUSolverTest {
@Test
fun testInvertOne() {
val matrix = MatrixContext.real.one(2, 2)
val inverted = MatrixContext.real.inverse(matrix)
val inverted = MatrixContext.real.inverseWithLUP(matrix)
assertEquals(matrix, inverted)
}
@ -37,7 +37,7 @@ class RealLUSolverTest {
1.0, 3.0
)
val inverted = MatrixContext.real.inverse(matrix)
val inverted = MatrixContext.real.inverseWithLUP(matrix)
val expected = Matrix.square(
0.375, -0.125,

2
kmath-coroutines/src/commonMain/kotlin/kscience/kmath/chains/Chain.kt
View File

@ -47,7 +47,7 @@ public fun <T> Iterator<T>.asChain(): Chain<T> = SimpleChain { next() }
public fun <T> Sequence<T>.asChain(): Chain<T> = iterator().asChain()
/**
* A simple chain of independent tokens
* A simple chain of independent tokens. [fork] returns the same chain.
*/
public class SimpleChain<out R>(private val gen: suspend () -> R) : Chain<R> {
public override suspend fun next(): R = gen()

4
kmath-dimensions/build.gradle.kts
View File

@ -18,3 +18,7 @@ kotlin.sourceSets {
}
}
}
readme{
maturity = ru.mipt.npm.gradle.Maturity.PROTOTYPE
}

4
kmath-dimensions/src/commonMain/kotlin/kscience/kmath/dimensions/Wrappers.kt
View File

@ -42,7 +42,7 @@ public interface DMatrix<T, R : Dimension, C : Dimension> : Structure2D<T> {
* An inline wrapper for a Matrix
*/
public inline class DMatrixWrapper<T, R : Dimension, C : Dimension>(
public val structure: Structure2D<T>
private val structure: Structure2D<T>
) : DMatrix<T, R, C> {
override val shape: IntArray get() = structure.shape
override operator fun get(i: Int, j: Int): T = structure[i, j]
@ -81,7 +81,7 @@ public inline class DPointWrapper<T, D : Dimension>(public val point: Point<T>)
/**
* Basic operations on dimension-safe matrices. Operates on [Matrix]
*/
public inline class DMatrixContext<T : Any, Ri : Ring<T>>(public val context: GenericMatrixContext<T, Ri>) {
public inline class DMatrixContext<T : Any, Ri : Ring<T>>(public val context: GenericMatrixContext<T, Ri, Matrix<T>>) {
public inline fun <reified R : Dimension, reified C : Dimension> Matrix<T>.coerce(): DMatrix<T, R, C> {
require(rowNum == Dimension.dim<R>().toInt()) {
"Row number mismatch: expected ${Dimension.dim<R>()} but found $rowNum"

24
kmath-ejml/src/main/kotlin/kscience/kmath/ejml/EjmlMatrixContext.kt
View File

@ -1,23 +1,22 @@
package kscience.kmath.ejml
import org.ejml.simple.SimpleMatrix
import kscience.kmath.linear.MatrixContext
import kscience.kmath.linear.Point
import kscience.kmath.operations.Space
import kscience.kmath.operations.invoke
import kscience.kmath.structures.Matrix
import org.ejml.simple.SimpleMatrix
/**
* Converts this matrix to EJML one.
*/
public fun Matrix<Double>.toEjml(): EjmlMatrix =
if (this is EjmlMatrix) this else EjmlMatrixContext.produce(rowNum, colNum) { i, j -> get(i, j) }
/**
* Represents context of basic operations operating with [EjmlMatrix].
*
* @author Iaroslav Postovalov
*/
public class EjmlMatrixContext(private val space: Space<Double>) : MatrixContext<Double> {
/**
* Converts this matrix to EJML one.
*/
public fun Matrix<Double>.toEjml(): EjmlMatrix =
if (this is EjmlMatrix) this else produce(rowNum, colNum) { i, j -> get(i, j) }
public object EjmlMatrixContext : MatrixContext<Double, EjmlMatrix> {
/**
* Converts this vector to EJML one.
@ -47,11 +46,10 @@ public class EjmlMatrixContext(private val space: Space<Double>) : MatrixContext
EjmlMatrix(toEjml().origin - b.toEjml().origin)
public override fun multiply(a: Matrix<Double>, k: Number): EjmlMatrix =
produce(a.rowNum, a.colNum) { i, j -> space { a[i, j] * k } }
produce(a.rowNum, a.colNum) { i, j -> a[i, j] * k.toDouble() }
public override operator fun Matrix<Double>.times(value: Double): EjmlMatrix = EjmlMatrix(toEjml().origin.scale(value))
public companion object
public override operator fun Matrix<Double>.times(value: Double): EjmlMatrix =
EjmlMatrix(toEjml().origin.scale(value))
}
/**

44
kmath-for-real/README.md Normal file
View File

@ -0,0 +1,44 @@
# Real number specialization module (`kmath-for-real`)
- [RealVector](src/commonMain/kotlin/kscience/kmath/real/RealVector.kt) : Numpy-like operations for Buffers/Points
- [RealMatrix](src/commonMain/kotlin/kscience/kmath/real/RealMatrix.kt) : Numpy-like operations for 2d real structures
- [grids](src/commonMain/kotlin/kscience/kmath/structures/grids.kt) : Uniform grid generators
> #### Artifact:
>
> This module artifact: `kscience.kmath:kmath-for-real:0.2.0-dev-4`.
>
> Bintray release version: [ ![Download](https://api.bintray.com/packages/mipt-npm/kscience/kmath-for-real/images/download.svg) ](https://bintray.com/mipt-npm/kscience/kmath-for-real/_latestVersion)
>
> Bintray development version: [ ![Download](https://api.bintray.com/packages/mipt-npm/dev/kmath-for-real/images/download.svg) ](https://bintray.com/mipt-npm/dev/kmath-for-real/_latestVersion)
>
> **Gradle:**
>
> ```gradle
> repositories {
> maven { url "https://dl.bintray.com/kotlin/kotlin-eap" }
> maven { url 'https://dl.bintray.com/mipt-npm/kscience' }
> maven { url 'https://dl.bintray.com/mipt-npm/dev' }
> maven { url 'https://dl.bintray.com/hotkeytlt/maven' }
> }
>
> dependencies {
> implementation 'kscience.kmath:kmath-for-real:0.2.0-dev-4'
> }
> ```
> **Gradle Kotlin DSL:**
>
> ```kotlin
> repositories {
> maven("https://dl.bintray.com/kotlin/kotlin-eap")
> maven("https://dl.bintray.com/mipt-npm/kscience")
> maven("https://dl.bintray.com/mipt-npm/dev")
> maven("https://dl.bintray.com/hotkeytlt/maven")
> }
>
> dependencies {
> implementation("kscience.kmath:kmath-for-real:0.2.0-dev-4")
> }
> ```

28
kmath-for-real/build.gradle.kts
View File

@ -7,3 +7,31 @@ kotlin.sourceSets.commonMain {
api(project(":kmath-core"))
}
}
readme {
description = """
Extension module that should be used to achieve numpy-like behavior.
All operations are specialized to work with `Double` numbers without declaring algebraic contexts.
One can still use generic algebras though.
""".trimIndent()
maturity = ru.mipt.npm.gradle.Maturity.EXPERIMENTAL
propertyByTemplate("artifact", rootProject.file("docs/templates/ARTIFACT-TEMPLATE.md"))
feature(
id = "RealVector",
description = "Numpy-like operations for Buffers/Points",
ref = "src/commonMain/kotlin/kscience/kmath/real/RealVector.kt"
)
feature(
id = "RealMatrix",
description = "Numpy-like operations for 2d real structures",
ref = "src/commonMain/kotlin/kscience/kmath/real/RealMatrix.kt"
)
feature(
id = "grids",
description = "Uniform grid generators",
ref = "src/commonMain/kotlin/kscience/kmath/structures/grids.kt"
)
}

5
kmath-for-real/docs/README-TEMPLATE.md Normal file
View File

@ -0,0 +1,5 @@
# Real number specialization module (`kmath-for-real`)
${features}
${artifact}

97
kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/RealMatrix.kt
View File

@ -1,12 +1,14 @@
package kscience.kmath.real
import kscience.kmath.linear.FeaturedMatrix
import kscience.kmath.linear.MatrixContext
import kscience.kmath.linear.RealMatrixContext.elementContext
import kscience.kmath.linear.VirtualMatrix
import kscience.kmath.linear.inverseWithLUP
import kscience.kmath.misc.UnstableKMathAPI
import kscience.kmath.operations.invoke
import kscience.kmath.operations.sum
import kscience.kmath.structures.Buffer
import kscience.kmath.structures.Matrix
import kscience.kmath.structures.RealBuffer
import kscience.kmath.structures.asIterable
import kotlin.math.pow
@ -23,7 +25,7 @@ import kotlin.math.pow
* Functions that help create a real (Double) matrix
*/
public typealias RealMatrix = Matrix<Double>
public typealias RealMatrix = FeaturedMatrix<Double>
public fun realMatrix(rowNum: Int, colNum: Int, initializer: (i: Int, j: Int) -> Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum, initializer)
@ -36,7 +38,7 @@ public fun Sequence<DoubleArray>.toMatrix(): RealMatrix = toList().let {
MatrixContext.real.produce(it.size, it[0].size) { row, col -> it[row][col] }
}
public fun Matrix<Double>.repeatStackVertical(n: Int): RealMatrix =
public fun RealMatrix.repeatStackVertical(n: Int): RealMatrix =
VirtualMatrix(rowNum * n, colNum) { row, col ->
get(if (row == 0) 0 else row % rowNum, col)
}
@ -45,76 +47,65 @@ public fun Matrix<Double>.repeatStackVertical(n: Int): RealMatrix =
* Operations for matrix and real number
*/
public operator fun Matrix<Double>.times(double: Double): RealMatrix =
public operator fun RealMatrix.times(double: Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col ->
this[row, col] * double
}
public operator fun Matrix<Double>.plus(double: Double): RealMatrix =
public operator fun RealMatrix.plus(double: Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col ->
this[row, col] + double
}
public operator fun Matrix<Double>.minus(double: Double): RealMatrix =
public operator fun RealMatrix.minus(double: Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col ->
this[row, col] - double
}
public operator fun Matrix<Double>.div(double: Double): RealMatrix =
public operator fun RealMatrix.div(double: Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col ->
this[row, col] / double
}
public operator fun Double.times(matrix: Matrix<Double>): RealMatrix =
public operator fun Double.times(matrix: RealMatrix): RealMatrix =
MatrixContext.real.produce(matrix.rowNum, matrix.colNum) { row, col ->
this * matrix[row, col]
}
public operator fun Double.plus(matrix: Matrix<Double>): RealMatrix =
public operator fun Double.plus(matrix: RealMatrix): RealMatrix =
MatrixContext.real.produce(matrix.rowNum, matrix.colNum) { row, col ->
this + matrix[row, col]
}
public operator fun Double.minus(matrix: Matrix<Double>): RealMatrix =
public operator fun Double.minus(matrix: RealMatrix): RealMatrix =
MatrixContext.real.produce(matrix.rowNum, matrix.colNum) { row, col ->
this - matrix[row, col]
}
// TODO: does this operation make sense? Should it be 'this/matrix[row, col]'?
//operator fun Double.div(matrix: Matrix<Double>) = MatrixContext.real.produce(matrix.rowNum, matrix.colNum) {
//operator fun Double.div(matrix: RealMatrix) = MatrixContext.real.produce(matrix.rowNum, matrix.colNum) {
// row, col -> matrix[row, col] / this
//}
/*
* Per-element (!) square and power operations
*/
public fun Matrix<Double>.square(): RealMatrix = MatrixContext.real.produce(rowNum, colNum) { row, col ->
this[row, col].pow(2)
}
public fun Matrix<Double>.pow(n: Int): RealMatrix = MatrixContext.real.produce(rowNum, colNum) { i, j ->
this[i, j].pow(n)
}
/*
* Operations on two matrices (per-element!)
*/
public operator fun Matrix<Double>.times(other: Matrix<Double>): RealMatrix =
@UnstableKMathAPI
public operator fun RealMatrix.times(other: RealMatrix): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col -> this[row, col] * other[row, col] }
public operator fun Matrix<Double>.plus(other: Matrix<Double>): RealMatrix =
public operator fun RealMatrix.plus(other: RealMatrix): RealMatrix =
MatrixContext.real.add(this, other)
public operator fun Matrix<Double>.minus(other: Matrix<Double>): RealMatrix =
public operator fun RealMatrix.minus(other: RealMatrix): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { row, col -> this[row, col] - other[row, col] }
/*
* Operations on columns
*/
public inline fun Matrix<Double>.appendColumn(crossinline mapper: (Buffer<Double>) -> Double): Matrix<Double> =
public inline fun RealMatrix.appendColumn(crossinline mapper: (Buffer<Double>) -> Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum + 1) { row, col ->
if (col < colNum)
this[row, col]
@ -122,28 +113,28 @@ public inline fun Matrix<Double>.appendColumn(crossinline mapper: (Buffer<Double
mapper(rows[row])
}
public fun Matrix<Double>.extractColumns(columnRange: IntRange): RealMatrix =
public fun RealMatrix.extractColumns(columnRange: IntRange): RealMatrix =
MatrixContext.real.produce(rowNum, columnRange.count()) { row, col ->
this[row, columnRange.first + col]
}
public fun Matrix<Double>.extractColumn(columnIndex: Int): RealMatrix =
public fun RealMatrix.extractColumn(columnIndex: Int): RealMatrix =
extractColumns(columnIndex..columnIndex)
public fun Matrix<Double>.sumByColumn(): RealBuffer = RealBuffer(colNum) { j ->
public fun RealMatrix.sumByColumn(): RealBuffer = RealBuffer(colNum) { j ->
val column = columns[j]
elementContext { sum(column.asIterable()) }
}
public fun Matrix<Double>.minByColumn(): RealBuffer = RealBuffer(colNum) { j ->
public fun RealMatrix.minByColumn(): RealBuffer = RealBuffer(colNum) { j ->
columns[j].asIterable().minOrNull() ?: error("Cannot produce min on empty column")
}
public fun Matrix<Double>.maxByColumn(): RealBuffer = RealBuffer(colNum) { j ->
public fun RealMatrix.maxByColumn(): RealBuffer = RealBuffer(colNum) { j ->
columns[j].asIterable().maxOrNull() ?: error("Cannot produce min on empty column")
}
public fun Matrix<Double>.averageByColumn(): RealBuffer = RealBuffer(colNum) { j ->
public fun RealMatrix.averageByColumn(): RealBuffer = RealBuffer(colNum) { j ->
columns[j].asIterable().average()
}
@ -151,7 +142,39 @@ public fun Matrix<Double>.averageByColumn(): RealBuffer = RealBuffer(colNum) { j
* Operations processing all elements
*/
public fun Matrix<Double>.sum(): Double = elements().map { (_, value) -> value }.sum()
public fun Matrix<Double>.min(): Double? = elements().map { (_, value) -> value }.minOrNull()
public fun Matrix<Double>.max(): Double? = elements().map { (_, value) -> value }.maxOrNull()
public fun Matrix<Double>.average(): Double = elements().map { (_, value) -> value }.average()
public fun RealMatrix.sum(): Double = elements().map { (_, value) -> value }.sum()
public fun RealMatrix.min(): Double? = elements().map { (_, value) -> value }.minOrNull()
public fun RealMatrix.max(): Double? = elements().map { (_, value) -> value }.maxOrNull()
public fun RealMatrix.average(): Double = elements().map { (_, value) -> value }.average()
public inline fun RealMatrix.map(transform: (Double) -> Double): RealMatrix =
MatrixContext.real.produce(rowNum, colNum) { i, j ->
transform(get(i, j))
}
/**
* Inverse a square real matrix using LUP decomposition
*/
public fun RealMatrix.inverseWithLUP(): RealMatrix = MatrixContext.real.inverseWithLUP(this)
//extended operations
public fun RealMatrix.pow(p: Double): RealMatrix = map { it.pow(p) }
public fun RealMatrix.pow(p: Int): RealMatrix = map { it.pow(p) }
public fun exp(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.exp(it) }
public fun sqrt(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.sqrt(it) }
public fun RealMatrix.square(): RealMatrix = map { it.pow(2) }
public fun sin(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.sin(it) }
public fun cos(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.cos(it) }
public fun tan(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.tan(it) }
public fun ln(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.ln(it) }
public fun log10(arg: RealMatrix): RealMatrix = arg.map { kotlin.math.log10(it) }

91
kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/RealVector.kt
View File

@ -1,47 +1,82 @@
package kscience.kmath.real
import kscience.kmath.linear.BufferVectorSpace
import kscience.kmath.linear.Point
import kscience.kmath.linear.VectorSpace
import kscience.kmath.operations.Norm
import kscience.kmath.operations.RealField
import kscience.kmath.operations.SpaceElement
import kscience.kmath.structures.Buffer
import kscience.kmath.structures.RealBuffer
import kscience.kmath.structures.asBuffer
import kscience.kmath.structures.asIterable
import kotlin.math.pow
import kotlin.math.sqrt
public typealias RealPoint = Point<Double>
public fun RealPoint.asVector(): RealVector = RealVector(this)
public fun DoubleArray.asVector(): RealVector = asBuffer().asVector()
public fun List<Double>.asVector(): RealVector = asBuffer().asVector()
public typealias RealVector = Point<Double>
public object VectorL2Norm : Norm<Point<out Number>, Double> {
override fun norm(arg: Point<out Number>): Double = sqrt(arg.asIterable().sumByDouble(Number::toDouble))
}
public inline class RealVector(private val point: Point<Double>) :
SpaceElement<RealPoint, RealVector, VectorSpace<Double, RealField>>, RealPoint {
public override val size: Int get() = point.size
public override val context: VectorSpace<Double, RealField> get() = space(point.size)
public operator fun Buffer.Companion.invoke(vararg doubles: Double): RealVector = doubles.asBuffer()
public override fun unwrap(): RealPoint = point
public override fun RealPoint.wrap(): RealVector = RealVector(this)
public override operator fun get(index: Int): Double = point[index]
public override operator fun iterator(): Iterator<Double> = point.iterator()
/**
* Fill the vector of given [size] with given [value]
*/
public fun Buffer.Companion.same(size: Int, value: Number): RealVector = real(size) { value.toDouble() }
public companion object {
private val spaceCache: MutableMap<Int, BufferVectorSpace<Double, RealField>> = hashMapOf()
// Transformation methods
public inline operator fun invoke(dim: Int, initializer: (Int) -> Double): RealVector =
RealVector(RealBuffer(dim, initializer))
public inline fun RealVector.map(transform: (Double) -> Double): RealVector =
Buffer.real(size) { transform(get(it)) }
public operator fun invoke(vararg values: Double): RealVector = values.asVector()
public inline fun RealVector.mapIndexed(transform: (index: Int, value: Double) -> Double): RealVector =
Buffer.real(size) { transform(it, get(it)) }
public fun space(dim: Int): BufferVectorSpace<Double, RealField> = spaceCache.getOrPut(dim) {
BufferVectorSpace(dim, RealField) { size, init -> Buffer.real(size, init) }
}
}
}
public operator fun RealVector.plus(other: RealVector): RealVector =
mapIndexed { index, value -> value + other[index] }
public operator fun RealVector.plus(number: Number): RealVector = map { it + number.toDouble() }
public operator fun Number.plus(vector: RealVector): RealVector = vector + this
public operator fun RealVector.unaryMinus(): Buffer<Double> = map { -it }
public operator fun RealVector.minus(other: RealVector): RealVector =
mapIndexed { index, value -> value - other[index] }
public operator fun RealVector.minus(number: Number): RealVector = map { it - number.toDouble() }
public operator fun Number.minus(vector: RealVector): RealVector = vector.map { toDouble() - it }
public operator fun RealVector.times(other: RealVector): RealVector =
mapIndexed { index, value -> value * other[index] }
public operator fun RealVector.times(number: Number): RealVector = map { it * number.toDouble() }
public operator fun Number.times(vector: RealVector): RealVector = vector * this
public operator fun RealVector.div(other: RealVector): RealVector =
mapIndexed { index, value -> value / other[index] }
public operator fun RealVector.div(number: Number): RealVector = map { it / number.toDouble() }
public operator fun Number.div(vector: RealVector): RealVector = vector.map { toDouble() / it }
//extended operations
public fun RealVector.pow(p: Double): RealVector = map { it.pow(p) }
public fun RealVector.pow(p: Int): RealVector = map { it.pow(p) }
public fun exp(vector: RealVector): RealVector = vector.map { kotlin.math.exp(it) }
public fun sqrt(vector: RealVector): RealVector = vector.map { kotlin.math.sqrt(it) }
public fun RealVector.square(): RealVector = map { it.pow(2) }
public fun sin(vector: RealVector): RealVector = vector.map { kotlin.math.sin(it) }
public fun cos(vector: RealVector): RealVector = vector.map { kotlin.math.cos(it) }
public fun tan(vector: RealVector): RealVector = vector.map { kotlin.math.tan(it) }
public fun ln(vector: RealVector): RealVector = vector.map { kotlin.math.ln(it) }
public fun log10(vector: RealVector): RealVector = vector.map { kotlin.math.log10(it) }

31
kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/dot.kt Normal file
View File

@ -0,0 +1,31 @@
package kscience.kmath.real
import kscience.kmath.linear.BufferMatrix
import kscience.kmath.structures.Buffer
import kscience.kmath.structures.RealBuffer
/**
* Optimized dot product for real matrices
*/
public infix fun BufferMatrix<Double>.dot(other: BufferMatrix<Double>): BufferMatrix<Double> {
require(colNum == other.rowNum) { "Matrix dot operation dimension mismatch: ($rowNum, $colNum) x (${other.rowNum}, ${other.colNum})" }
val resultArray = DoubleArray(this.rowNum * other.colNum)
//convert to array to insure there is no memory indirection
fun Buffer<out Double>.unsafeArray() = if (this is RealBuffer)
this.array
else
DoubleArray(size) { get(it) }
val a = this.buffer.unsafeArray()
val b = other.buffer.unsafeArray()
for (i in (0 until rowNum))
for (j in (0 until other.colNum))
for (k in (0 until colNum))
resultArray[i * other.colNum + j] += a[i * colNum + k] * b[k * other.colNum + j]
val buffer = RealBuffer(resultArray)
return BufferMatrix(rowNum, other.colNum, buffer)
}

3
kmath-for-real/src/commonMain/kotlin/kscience/kmath/real/grids.kt
View File

@ -1,6 +1,5 @@
package kscience.kmath.real
import kscience.kmath.linear.Point
import kscience.kmath.structures.asBuffer
import kotlin.math.abs
@ -34,7 +33,7 @@ public fun ClosedFloatingPointRange<Double>.toSequenceWithStep(step: Double): Se
}
}
public infix fun ClosedFloatingPointRange<Double>.step(step: Double): Point<Double> =
public infix fun ClosedFloatingPointRange<Double>.step(step: Double): RealVector =
toSequenceWithStep(step).toList().asBuffer()
/**

16
kmath-for-real/src/commonTest/kotlin/kaceince/kmath/real/RealVectorTest.kt
View File

@ -1,33 +1,33 @@
package kaceince.kmath.real
import kscience.kmath.linear.MatrixContext
import kscience.kmath.linear.asMatrix
import kscience.kmath.linear.transpose
import kscience.kmath.linear.*
import kscience.kmath.operations.invoke
import kscience.kmath.real.RealVector
import kscience.kmath.real.plus
import kscience.kmath.structures.Buffer
import kotlin.test.Test
import kotlin.test.assertEquals
internal class RealVectorTest {
@Test
fun testSum() {
val vector1 = RealVector(5) { it.toDouble() }
val vector2 = RealVector(5) { 5 - it.toDouble() }
val vector1 = Buffer.real(5) { it.toDouble() }
val vector2 = Buffer.real(5) { 5 - it.toDouble() }
val sum = vector1 + vector2
assertEquals(5.0, sum[2])
}
@Test
fun testVectorToMatrix() {
val vector = RealVector(5) { it.toDouble() }
val vector = Buffer.real(5) { it.toDouble() }
val matrix = vector.asMatrix()
assertEquals(4.0, matrix[4, 0])
}
@Test
fun testDot() {
val vector1 = RealVector(5) { it.toDouble() }
val vector2 = RealVector(5) { 5 - it.toDouble() }
val vector1 = Buffer.real(5) { it.toDouble() }
val vector2 = Buffer.real(5) { 5 - it.toDouble() }
val matrix1 = vector1.asMatrix()
val matrix2 = vector2.asMatrix().transpose()
val product = MatrixContext.real { matrix1 dot matrix2 }

5
kmath-histograms/src/commonMain/kotlin/kscience/kmath/histogram/RealHistogram.kt
View File

@ -3,7 +3,6 @@ package kscience.kmath.histogram
import kscience.kmath.linear.Point
import kscience.kmath.operations.SpaceOperations
import kscience.kmath.operations.invoke
import kscience.kmath.real.asVector
import kscience.kmath.structures.*
import kotlin.math.floor
@ -123,8 +122,8 @@ public class RealHistogram(
*```
*/
public fun fromRanges(vararg ranges: ClosedFloatingPointRange<Double>): RealHistogram = RealHistogram(
ranges.map(ClosedFloatingPointRange<Double>::start).asVector(),
ranges.map(ClosedFloatingPointRange<Double>::endInclusive).asVector()
ranges.map(ClosedFloatingPointRange<Double>::start).asBuffer(),
ranges.map(ClosedFloatingPointRange<Double>::endInclusive).asBuffer()
)
/**

2
kmath-histograms/src/commonTest/kotlin/scietifik/kmath/histogram/MultivariateHistogramTest.kt
View File

@ -4,6 +4,8 @@ import kscience.kmath.histogram.RealHistogram
import kscience.kmath.histogram.fill
import kscience.kmath.histogram.put
import kscience.kmath.real.RealVector
import kscience.kmath.real.invoke
import kscience.kmath.structures.Buffer
import kotlin.random.Random
import kotlin.test.*

4
kmath-histograms/src/jvmMain/kotlin/kscience/kmath/histogram/UnivariateHistogram.kt
View File

@ -1,8 +1,8 @@
package kscience.kmath.histogram
import kscience.kmath.real.RealVector
import kscience.kmath.real.asVector
import kscience.kmath.structures.Buffer
import kscience.kmath.structures.asBuffer
import java.util.*
import kotlin.math.floor
@ -16,7 +16,7 @@ public class UnivariateBin(
//TODO add weighting
public override val value: Number get() = counter.sum()
public override val center: RealVector get() = doubleArrayOf(position).asVector()
public override val center: RealVector get() = doubleArrayOf(position).asBuffer()
public override val dimension: Int get() = 1
public operator fun contains(value: Double): Boolean = value in (position - size / 2)..(position + size / 2)

4
kmath-kotlingrad/build.gradle.kts
View File

@ -3,7 +3,7 @@ plugins {
}
dependencies {
implementation("com.github.breandan:kaliningraph:0.1.2")
implementation("com.github.breandan:kotlingrad:0.3.7")
implementation("com.github.breandan:kaliningraph:0.1.4")
implementation("com.github.breandan:kotlingrad:0.4.0")
api(project(":kmath-ast"))
}

2
kmath-kotlingrad/src/main/kotlin/kscience/kmath/kotlingrad/DifferentiableMstExpression.kt
View File

@ -1,6 +1,6 @@
package kscience.kmath.kotlingrad
import edu.umontreal.kotlingrad.experimental.SFun
import edu.umontreal.kotlingrad.api.SFun
import kscience.kmath.ast.MST
import kscience.kmath.ast.MstAlgebra
import kscience.kmath.ast.MstExpression

4
kmath-kotlingrad/src/main/kotlin/kscience/kmath/kotlingrad/KMathNumber.kt
View File

@ -1,7 +1,7 @@
package kscience.kmath.kotlingrad
import edu.umontreal.kotlingrad.experimental.RealNumber
import edu.umontreal.kotlingrad.experimental.SConst
import edu.umontreal.kotlingrad.api.RealNumber
import edu.umontreal.kotlingrad.api.SConst
import kscience.kmath.operations.NumericAlgebra
/**

2
kmath-kotlingrad/src/main/kotlin/kscience/kmath/kotlingrad/ScalarsAdapters.kt
View File

@ -1,6 +1,6 @@
package kscience.kmath.kotlingrad
import edu.umontreal.kotlingrad.experimental.*
import edu.umontreal.kotlingrad.api.*
import kscience.kmath.ast.MST
import kscience.kmath.ast.MstAlgebra
import kscience.kmath.ast.MstExtendedField

2
kmath-kotlingrad/src/test/kotlin/kscience/kmath/kotlingrad/AdaptingTests.kt
View File

@ -1,6 +1,6 @@
package kscience.kmath.kotlingrad
import edu.umontreal.kotlingrad.experimental.*
import edu.umontreal.kotlingrad.api.*
import kscience.kmath.asm.compile
import kscience.kmath.ast.MstAlgebra
import kscience.kmath.ast.MstExpression

8
kmath-nd4j/README.md
View File

@ -2,14 +2,14 @@
This subproject implements the following features:
- [nd4jarraystrucure](src/commonMain/kotlin/kscience/kmath/operations/Algebra.kt) : NDStructure wrapper for INDArray
- [nd4jarraystructure](src/commonMain/kotlin/kscience/kmath/operations/Algebra.kt) : NDStructure wrapper for INDArray
- [nd4jarrayrings](src/commonMain/kotlin/kscience/kmath/structures/NDStructure.kt) : Rings over Nd4jArrayStructure of Int and Long
- [nd4jarrayfields](src/commonMain/kotlin/kscience/kmath/structures/Buffers.kt) : Fields over Nd4jArrayStructure of Float and Double
> #### Artifact:
>
> This module artifact: `kscience.kmath:kmath-nd4j:0.2.0-dev-3`.
> This module artifact: `kscience.kmath:kmath-nd4j:0.2.0-dev-4`.
>
> Bintray release version: [ ![Download](https://api.bintray.com/packages/mipt-npm/kscience/kmath-nd4j/images/download.svg) ](https://bintray.com/mipt-npm/kscience/kmath-nd4j/_latestVersion)
>
@ -27,7 +27,7 @@ This subproject implements the following features:
> }
>
> dependencies {
> implementation 'kscience.kmath:kmath-nd4j:0.2.0-dev-3'
> implementation 'kscience.kmath:kmath-nd4j:0.2.0-dev-4'
> }
> ```
> **Gradle Kotlin DSL:**
@ -41,7 +41,7 @@ This subproject implements the following features:
> }
>
> dependencies {
> implementation("kscience.kmath:kmath-nd4j:0.2.0-dev-3")
> implementation("kscience.kmath:kmath-nd4j:0.2.0-dev-4")
> }
> ```