diff --git a/CHANGELOG.md b/CHANGELOG.md
index 4c6d94a2c..2388cd159 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -7,12 +7,13 @@
- Better trigonometric and hyperbolic functions for `AutoDiffField` (https://github.com/mipt-npm/kmath/pull/140).
- Automatic README generation for features (#139)
- Native support for `memory`, `core` and `dimensions`
-- `kmath-ejml` to supply EJML SimpleMatrix wrapper.
+- `kmath-ejml` to supply EJML SimpleMatrix wrapper (https://github.com/mipt-npm/kmath/pull/136).
- A separate `Symbol` entity, which is used for global unbound symbol.
- A `Symbol` indexing scope.
- Basic optimization API for Commons-math.
- Chi squared optimization for array-like data in CM
- `Fitting` utility object in prob/stat
+- ND4J support module submitting `NDStructure` and `NDAlgebra` over `INDArray`.
- Basic Quaternion vector support.
### Changed
diff --git a/README.md b/README.md
index afab32dcf..c4e3e5374 100644
--- a/README.md
+++ b/README.md
@@ -8,41 +8,50 @@ Bintray: [ ![Download](https://api.bintray.com/packages/mipt-npm/kscience
Bintray-dev: [ ![Download](https://api.bintray.com/packages/mipt-npm/dev/kmath-core/images/download.svg) ](https://bintray.com/mipt-npm/dev/kmath-core/_latestVersion)
# KMath
-Could be pronounced as `key-math`.
-The Kotlin MATHematics library was initially intended as a Kotlin-based analog to Python's `numpy` library. Later we found that kotlin is much more flexible language and allows superior architecture designs. In contrast to `numpy` and `scipy` it is modular and has a lightweight core. The `numpy`-like experience could be achieved with [kmath-for-real](/kmath-for-real) extension module.
+
+Could be pronounced as `key-math`. The Kotlin MATHematics library was initially intended as a Kotlin-based analog to
+Python's NumPy library. Later we found that kotlin is much more flexible language and allows superior architecture
+designs. In contrast to `numpy` and `scipy` it is modular and has a lightweight core. The `numpy`-like experience could
+be achieved with [kmath-for-real](/kmath-for-real) extension module.
## Publications and talks
+
* [A conceptual article about context-oriented design](https://proandroiddev.com/an-introduction-context-oriented-programming-in-kotlin-2e79d316b0a2)
* [Another article about context-oriented design](https://proandroiddev.com/diving-deeper-into-context-oriented-programming-in-kotlin-3ecb4ec38814)
* [ACAT 2019 conference paper](https://aip.scitation.org/doi/abs/10.1063/1.5130103)
# Goal
-* Provide a flexible and powerful API to work with mathematics abstractions in Kotlin-multiplatform (JVM and JS for now and Native in future).
+
+* Provide a flexible and powerful API to work with mathematics abstractions in Kotlin-multiplatform (JVM, JS and Native).
* Provide basic multiplatform implementations for those abstractions (without significant performance optimization).
* Provide bindings and wrappers with those abstractions for popular optimized platform libraries.
## Non-goals
-* Be like Numpy. It was the idea at the beginning, but we decided that we can do better in terms of API.
-* Provide best performance out of the box. We have specialized libraries for that. Need only API wrappers for them.
+
+* Be like NumPy. It was the idea at the beginning, but we decided that we can do better in terms of API.
+* Provide the best performance out of the box. We have specialized libraries for that. Need only API wrappers for them.
* Cover all cases as immediately and in one bundle. We will modularize everything and add new features gradually.
-* Provide specialized behavior in the core. API is made generic on purpose, so one needs to specialize for types, like for `Double` in the core. For that we will have specialization modules like `for-real`, which will give better experience for those, who want to work with specific types.
+* Provide specialized behavior in the core. API is made generic on purpose, so one needs to specialize for types, like
+for `Double` in the core. For that we will have specialization modules like `for-real`, which will give better
+experience for those, who want to work with specific types.
## Features
-Actual feature list is [here](/docs/features.md)
+Current feature list is [here](/docs/features.md)
* **Algebra**
- * Algebraic structures like rings, spaces and field (**TODO** add example to wiki)
+ * Algebraic structures like rings, spaces and fields (**TODO** add example to wiki)
* Basic linear algebra operations (sums, products, etc.), backed by the `Space` API.
- * Complex numbers backed by the `Field` API (meaning that they will be usable in any structure like vectors and N-dimensional arrays).
+ * Complex numbers backed by the `Field` API (meaning they will be usable in any structure like vectors and
+ N-dimensional arrays).
* Advanced linear algebra operations like matrix inversion and LU decomposition.
* **Array-like structures** Full support of many-dimensional array-like structures
including mixed arithmetic operations and function operations over arrays and numbers (with the added benefit of static type checking).
-* **Expressions** By writing a single mathematical expression
-once, users will be able to apply different types of objects to the expression by providing a context. Expressions
-can be used for a wide variety of purposes from high performance calculations to code generation.
+* **Expressions** By writing a single mathematical expression once, users will be able to apply different types of
+objects to the expression by providing a context. Expressions can be used for a wide variety of purposes from high
+performance calculations to code generation.
* **Histograms** Fast multi-dimensional histograms.
@@ -50,9 +59,10 @@ can be used for a wide variety of purposes from high performance calculations to
* **Type-safe dimensions** Type-safe dimensions for matrix operations.
-* **Commons-math wrapper** It is planned to gradually wrap most parts of [Apache commons-math](http://commons.apache.org/proper/commons-math/)
- library in Kotlin code and maybe rewrite some parts to better suit the Kotlin programming paradigm, however there is no fixed roadmap for that. Feel free
- to submit a feature request if you want something to be done first.
+* **Commons-math wrapper** It is planned to gradually wrap most parts of
+[Apache commons-math](http://commons.apache.org/proper/commons-math/) library in Kotlin code and maybe rewrite some
+parts to better suit the Kotlin programming paradigm, however there is no established roadmap for that. Feel free to
+submit a feature request if you want something to be implemented first.
## Planned features
@@ -151,6 +161,18 @@ can be used for a wide variety of purposes from high performance calculations to
> **Maturity**: EXPERIMENTAL
+* ### [kmath-nd4j](kmath-nd4j)
+> ND4J NDStructure implementation and according NDAlgebra classes
+>
+> **Maturity**: EXPERIMENTAL
+>
+> **Features:**
+> - [nd4jarraystrucure](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
+
+
+
* ### [kmath-stat](kmath-stat)
>
>
@@ -166,39 +188,69 @@ can be used for a wide variety of purposes from high performance calculations to
## Multi-platform support
-KMath is developed as a multi-platform library, which means that most of the interfaces are declared in the [common module](/kmath-core/src/commonMain). Implementation is also done in the common module wherever possible. In some cases, features are delegated to platform-specific implementations even if they could be done in the common module for performance reasons. Currently, the JVM is the main focus of development, however Kotlin/Native and Kotlin/JS contributions are also welcome.
+KMath is developed as a multi-platform library, which means that most of the interfaces are declared in the
+[common source sets](/kmath-core/src/commonMain) and implemented there wherever it is possible. In some cases, features
+are delegated to platform-specific implementations even if they could be provided in the common module for performance
+reasons. Currently, the Kotlin/JVM is the primary platform, however Kotlin/Native and Kotlin/JS contributions and
+feedback are also welcome.
## Performance
-Calculation performance is one of major goals of KMath in the future, but in some cases it is not possible to achieve both performance and flexibility. We expect to focus on creating convenient universal API first and then work on increasing performance for specific cases. We expect the worst KMath benchmarks will perform better than native Python, but worse than optimized native/SciPy (mostly due to boxing operations on primitive numbers). The best performance of optimized parts could be better than SciPy.
+Calculation performance is one of major goals of KMath in the future, but in some cases it is impossible to achieve
+both performance and flexibility.
-### Dependency
+We expect to focus on creating convenient universal API first and then work on increasing performance for specific
+cases. We expect the worst KMath benchmarks will perform better than native Python, but worse than optimized
+native/SciPy (mostly due to boxing operations on primitive numbers). The best performance of optimized parts could be
+better than SciPy.
-Release artifacts are accessible from bintray with following configuration (see documentation for [kotlin-multiplatform](https://kotlinlang.org/docs/reference/multiplatform.html) form more details):
+### Repositories
+
+Release artifacts are accessible from bintray with following configuration (see documentation of
+[Kotlin Multiplatform](https://kotlinlang.org/docs/reference/multiplatform.html) for more details):
```kotlin
-repositories{
+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-2")
- //api("kscience.kmath:kmath-core-jvm:0.2.0-dev-2") for jvm-specific version
+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
}
```
Gradle `6.0+` is required for multiplatform artifacts.
-### Development
+#### Development
+
+Development builds are uploaded to the separate repository:
-Development builds are accessible from the reposirtory
```kotlin
-repositories{
+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()
}
```
-with the same artifact names.
## Contributing
-The project requires a lot of additional work. The most important thing we need is a feedback about what features are required the most. Feel free to open feature issues with requests. We are also welcome to code contributions, especially in issues marked as [waiting for a hero](https://github.com/mipt-npm/kmath/labels/waiting%20for%20a%20hero).
\ No newline at end of file
+The project requires a lot of additional work. The most important thing we need is a feedback about what features are
+required the most. Feel free to create feature requests. We are also welcome to code contributions,
+especially in issues marked with
+[waiting for a hero](https://github.com/mipt-npm/kmath/labels/waiting%20for%20a%20hero) label.
diff --git a/build.gradle.kts b/build.gradle.kts
index acb9f3b68..3514c91e6 100644
--- a/build.gradle.kts
+++ b/build.gradle.kts
@@ -1,17 +1,26 @@
+import ru.mipt.npm.gradle.KSciencePublishPlugin
+
plugins {
id("ru.mipt.npm.project")
}
-val kmathVersion: String by extra("0.2.0-dev-3")
-val bintrayRepo: String by extra("kscience")
-val githubProject: String by extra("kmath")
+internal val kmathVersion: String by extra("0.2.0-dev-3")
+internal val bintrayRepo: String by extra("kscience")
+internal val githubProject: String by extra("kmath")
allprojects {
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/hotkeytlt/maven")
+ maven("https://dl.bintray.com/mipt-npm/dev")
+ maven("https://dl.bintray.com/mipt-npm/kscience")
+ maven("https://jitpack.io")
+ maven("http://logicrunch.research.it.uu.se/maven/")
+ mavenCentral()
}
group = "kscience.kmath"
@@ -19,13 +28,13 @@ allprojects {
}
subprojects {
- if (name.startsWith("kmath")) apply()
+ if (name.startsWith("kmath")) apply()
}
readme {
readmeTemplate = file("docs/templates/README-TEMPLATE.md")
}
-apiValidation{
+apiValidation {
validationDisabled = true
-}
\ No newline at end of file
+}
diff --git a/docs/templates/README-TEMPLATE.md b/docs/templates/README-TEMPLATE.md
index 5117e0694..ee1df818c 100644
--- a/docs/templates/README-TEMPLATE.md
+++ b/docs/templates/README-TEMPLATE.md
@@ -8,41 +8,50 @@ Bintray: [ ![Download](https://api.bintray.com/packages/mipt-npm/kscience
Bintray-dev: [ ![Download](https://api.bintray.com/packages/mipt-npm/dev/kmath-core/images/download.svg) ](https://bintray.com/mipt-npm/dev/kmath-core/_latestVersion)
# KMath
-Could be pronounced as `key-math`.
-The Kotlin MATHematics library was initially intended as a Kotlin-based analog to Python's `numpy` library. Later we found that kotlin is much more flexible language and allows superior architecture designs. In contrast to `numpy` and `scipy` it is modular and has a lightweight core. The `numpy`-like experience could be achieved with [kmath-for-real](/kmath-for-real) extension module.
+
+Could be pronounced as `key-math`. The Kotlin MATHematics library was initially intended as a Kotlin-based analog to
+Python's NumPy library. Later we found that kotlin is much more flexible language and allows superior architecture
+designs. In contrast to `numpy` and `scipy` it is modular and has a lightweight core. The `numpy`-like experience could
+be achieved with [kmath-for-real](/kmath-for-real) extension module.
## Publications and talks
+
* [A conceptual article about context-oriented design](https://proandroiddev.com/an-introduction-context-oriented-programming-in-kotlin-2e79d316b0a2)
* [Another article about context-oriented design](https://proandroiddev.com/diving-deeper-into-context-oriented-programming-in-kotlin-3ecb4ec38814)
* [ACAT 2019 conference paper](https://aip.scitation.org/doi/abs/10.1063/1.5130103)
# Goal
-* Provide a flexible and powerful API to work with mathematics abstractions in Kotlin-multiplatform (JVM and JS for now and Native in future).
+
+* Provide a flexible and powerful API to work with mathematics abstractions in Kotlin-multiplatform (JVM, JS and Native).
* Provide basic multiplatform implementations for those abstractions (without significant performance optimization).
* Provide bindings and wrappers with those abstractions for popular optimized platform libraries.
## Non-goals
-* Be like Numpy. It was the idea at the beginning, but we decided that we can do better in terms of API.
-* Provide best performance out of the box. We have specialized libraries for that. Need only API wrappers for them.
+
+* Be like NumPy. It was the idea at the beginning, but we decided that we can do better in terms of API.
+* Provide the best performance out of the box. We have specialized libraries for that. Need only API wrappers for them.
* Cover all cases as immediately and in one bundle. We will modularize everything and add new features gradually.
-* Provide specialized behavior in the core. API is made generic on purpose, so one needs to specialize for types, like for `Double` in the core. For that we will have specialization modules like `for-real`, which will give better experience for those, who want to work with specific types.
+* Provide specialized behavior in the core. API is made generic on purpose, so one needs to specialize for types, like
+for `Double` in the core. For that we will have specialization modules like `for-real`, which will give better
+experience for those, who want to work with specific types.
## Features
-Actual feature list is [here](/docs/features.md)
+Current feature list is [here](/docs/features.md)
* **Algebra**
- * Algebraic structures like rings, spaces and field (**TODO** add example to wiki)
+ * Algebraic structures like rings, spaces and fields (**TODO** add example to wiki)
* Basic linear algebra operations (sums, products, etc.), backed by the `Space` API.
- * Complex numbers backed by the `Field` API (meaning that they will be usable in any structure like vectors and N-dimensional arrays).
+ * Complex numbers backed by the `Field` API (meaning they will be usable in any structure like vectors and
+ N-dimensional arrays).
* Advanced linear algebra operations like matrix inversion and LU decomposition.
* **Array-like structures** Full support of many-dimensional array-like structures
including mixed arithmetic operations and function operations over arrays and numbers (with the added benefit of static type checking).
-* **Expressions** By writing a single mathematical expression
-once, users will be able to apply different types of objects to the expression by providing a context. Expressions
-can be used for a wide variety of purposes from high performance calculations to code generation.
+* **Expressions** By writing a single mathematical expression once, users will be able to apply different types of
+objects to the expression by providing a context. Expressions can be used for a wide variety of purposes from high
+performance calculations to code generation.
* **Histograms** Fast multi-dimensional histograms.
@@ -50,9 +59,10 @@ can be used for a wide variety of purposes from high performance calculations to
* **Type-safe dimensions** Type-safe dimensions for matrix operations.
-* **Commons-math wrapper** It is planned to gradually wrap most parts of [Apache commons-math](http://commons.apache.org/proper/commons-math/)
- library in Kotlin code and maybe rewrite some parts to better suit the Kotlin programming paradigm, however there is no fixed roadmap for that. Feel free
- to submit a feature request if you want something to be done first.
+* **Commons-math wrapper** It is planned to gradually wrap most parts of
+[Apache commons-math](http://commons.apache.org/proper/commons-math/) library in Kotlin code and maybe rewrite some
+parts to better suit the Kotlin programming paradigm, however there is no established roadmap for that. Feel free to
+submit a feature request if you want something to be implemented first.
## Planned features
@@ -72,39 +82,53 @@ $modules
## Multi-platform support
-KMath is developed as a multi-platform library, which means that most of the interfaces are declared in the [common module](/kmath-core/src/commonMain). Implementation is also done in the common module wherever possible. In some cases, features are delegated to platform-specific implementations even if they could be done in the common module for performance reasons. Currently, the JVM is the main focus of development, however Kotlin/Native and Kotlin/JS contributions are also welcome.
+KMath is developed as a multi-platform library, which means that most of the interfaces are declared in the
+[common source sets](/kmath-core/src/commonMain) and implemented there wherever it is possible. In some cases, features
+are delegated to platform-specific implementations even if they could be provided in the common module for performance
+reasons. Currently, the Kotlin/JVM is the primary platform, however Kotlin/Native and Kotlin/JS contributions and
+feedback are also welcome.
## Performance
-Calculation performance is one of major goals of KMath in the future, but in some cases it is not possible to achieve both performance and flexibility. We expect to focus on creating convenient universal API first and then work on increasing performance for specific cases. We expect the worst KMath benchmarks will perform better than native Python, but worse than optimized native/SciPy (mostly due to boxing operations on primitive numbers). The best performance of optimized parts could be better than SciPy.
+Calculation performance is one of major goals of KMath in the future, but in some cases it is impossible to achieve
+both performance and flexibility.
-### Dependency
+We expect to focus on creating convenient universal API first and then work on increasing performance for specific
+cases. We expect the worst KMath benchmarks will perform better than native Python, but worse than optimized
+native/SciPy (mostly due to boxing operations on primitive numbers). The best performance of optimized parts could be
+better than SciPy.
-Release artifacts are accessible from bintray with following configuration (see documentation for [kotlin-multiplatform](https://kotlinlang.org/docs/reference/multiplatform.html) form more details):
+### Repositories
+
+Release artifacts are accessible from bintray with following configuration (see documentation of
+[Kotlin Multiplatform](https://kotlinlang.org/docs/reference/multiplatform.html) for more details):
```kotlin
-repositories{
+repositories {
maven("https://dl.bintray.com/mipt-npm/kscience")
}
-dependencies{
+dependencies {
api("kscience.kmath:kmath-core:$version")
- //api("kscience.kmath:kmath-core-jvm:$version") for jvm-specific version
+ // api("kscience.kmath:kmath-core-jvm:$version") for jvm-specific version
}
```
Gradle `6.0+` is required for multiplatform artifacts.
-### Development
+#### Development
+
+Development builds are uploaded to the separate repository:
-Development builds are accessible from the reposirtory
```kotlin
-repositories{
+repositories {
maven("https://dl.bintray.com/mipt-npm/dev")
}
```
-with the same artifact names.
## Contributing
-The project requires a lot of additional work. The most important thing we need is a feedback about what features are required the most. Feel free to open feature issues with requests. We are also welcome to code contributions, especially in issues marked as [waiting for a hero](https://github.com/mipt-npm/kmath/labels/waiting%20for%20a%20hero).
\ No newline at end of file
+The project requires a lot of additional work. The most important thing we need is a feedback about what features are
+required the most. Feel free to create feature requests. We are also welcome to code contributions,
+especially in issues marked with
+[waiting for a hero](https://github.com/mipt-npm/kmath/labels/waiting%20for%20a%20hero) label.
diff --git a/examples/build.gradle.kts b/examples/build.gradle.kts
index 0dcc58f79..d42627ff0 100644
--- a/examples/build.gradle.kts
+++ b/examples/build.gradle.kts
@@ -8,18 +8,25 @@ plugins {
}
allOpen.annotation("org.openjdk.jmh.annotations.State")
+sourceSets.register("benchmarks")
repositories {
- maven("https://dl.bintray.com/mipt-npm/kscience")
+ 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://dl.bintray.com/kotlin/kotlin-dev/")
+ maven("https://dl.bintray.com/mipt-npm/kscience")
+ maven("https://jitpack.io")
+ maven("http://logicrunch.research.it.uu.se/maven/")
mavenCentral()
}
-sourceSets.register("benchmarks")
-
dependencies {
implementation(project(":kmath-ast"))
+ implementation(project(":kmath-kotlingrad"))
implementation(project(":kmath-core"))
implementation(project(":kmath-coroutines"))
implementation(project(":kmath-commons"))
@@ -27,6 +34,20 @@ dependencies {
implementation(project(":kmath-viktor"))
implementation(project(":kmath-dimensions"))
implementation(project(":kmath-ejml"))
+ implementation(project(":kmath-nd4j"))
+ implementation("org.deeplearning4j:deeplearning4j-core:1.0.0-beta7")
+ implementation("org.nd4j:nd4j-native:1.0.0-beta7")
+
+// uncomment if your system supports AVX2
+// val os = System.getProperty("os.name")
+//
+// if (System.getProperty("os.arch") in arrayOf("x86_64", "amd64")) when {
+// os.startsWith("Windows") -> implementation("org.nd4j:nd4j-native:1.0.0-beta7:windows-x86_64-avx2")
+// os == "Linux" -> implementation("org.nd4j:nd4j-native:1.0.0-beta7:linux-x86_64-avx2")
+// os == "Mac OS X" -> implementation("org.nd4j:nd4j-native:1.0.0-beta7:macosx-x86_64-avx2")
+// } else
+ implementation("org.nd4j:nd4j-native-platform:1.0.0-beta7")
+
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")
@@ -55,4 +76,6 @@ kotlin.sourceSets.all {
}
}
-tasks.withType { kotlinOptions.jvmTarget = "11" }
+tasks.withType {
+ kotlinOptions.jvmTarget = "11"
+}
diff --git a/examples/src/main/kotlin/kscience/kmath/ast/ExpressionsInterpretersBenchmark.kt b/examples/src/main/kotlin/kscience/kmath/ast/ExpressionsInterpretersBenchmark.kt
index 3279d72ce..a4806ed68 100644
--- a/examples/src/main/kotlin/kscience/kmath/ast/ExpressionsInterpretersBenchmark.kt
+++ b/examples/src/main/kotlin/kscience/kmath/ast/ExpressionsInterpretersBenchmark.kt
@@ -9,11 +9,11 @@ import kscience.kmath.operations.RealField
import kotlin.random.Random
import kotlin.system.measureTimeMillis
-class ExpressionsInterpretersBenchmark {
+internal class ExpressionsInterpretersBenchmark {
private val algebra: Field = RealField
fun functionalExpression() {
val expr = algebra.expressionInField {
- variable("x") * const(2.0) + const(2.0) / variable("x") - const(16.0)
+ symbol("x") * const(2.0) + const(2.0) / symbol("x") - const(16.0)
}
invokeAndSum(expr)
@@ -47,6 +47,16 @@ class ExpressionsInterpretersBenchmark {
}
}
+/**
+ * This benchmark compares basically evaluation of simple function with MstExpression interpreter, ASM backend and
+ * core FunctionalExpressions API.
+ *
+ * The expected rating is:
+ *
+ * 1. ASM.
+ * 2. MST.
+ * 3. FE.
+ */
fun main() {
val benchmark = ExpressionsInterpretersBenchmark()
diff --git a/examples/src/main/kotlin/kscience/kmath/ast/KotlingradSupport.kt b/examples/src/main/kotlin/kscience/kmath/ast/KotlingradSupport.kt
new file mode 100644
index 000000000..b3c827503
--- /dev/null
+++ b/examples/src/main/kotlin/kscience/kmath/ast/KotlingradSupport.kt
@@ -0,0 +1,24 @@
+package kscience.kmath.ast
+
+import kscience.kmath.asm.compile
+import kscience.kmath.expressions.derivative
+import kscience.kmath.expressions.invoke
+import kscience.kmath.expressions.symbol
+import kscience.kmath.kotlingrad.differentiable
+import kscience.kmath.operations.RealField
+
+/**
+ * In this example, x^2-4*x-44 function is differentiated with Kotlin∇, and the autodiff result is compared with
+ * valid derivative.
+ */
+fun main() {
+ val x by symbol
+
+ val actualDerivative = MstExpression(RealField, "x^2-4*x-44".parseMath())
+ .differentiable()
+ .derivative(x)
+ .compile()
+
+ val expectedDerivative = MstExpression(RealField, "2*x-4".parseMath()).compile()
+ assert(actualDerivative("x" to 123.0) == expectedDerivative("x" to 123.0))
+}
diff --git a/examples/src/main/kotlin/kscience/kmath/structures/NDField.kt b/examples/src/main/kotlin/kscience/kmath/structures/NDField.kt
index 28bfab779..e53af0dee 100644
--- a/examples/src/main/kotlin/kscience/kmath/structures/NDField.kt
+++ b/examples/src/main/kotlin/kscience/kmath/structures/NDField.kt
@@ -1,8 +1,10 @@
package kscience.kmath.structures
import kotlinx.coroutines.GlobalScope
+import kscience.kmath.nd4j.Nd4jArrayField
import kscience.kmath.operations.RealField
import kscience.kmath.operations.invoke
+import org.nd4j.linalg.factory.Nd4j
import kotlin.contracts.InvocationKind
import kotlin.contracts.contract
import kotlin.system.measureTimeMillis
@@ -14,6 +16,8 @@ internal inline fun measureAndPrint(title: String, block: () -> Unit) {
}
fun main() {
+ // initializing Nd4j
+ Nd4j.zeros(0)
val dim = 1000
val n = 1000
@@ -23,6 +27,8 @@ fun main() {
val specializedField = NDField.real(dim, dim)
//A generic boxing field. It should be used for objects, not primitives.
val genericField = NDField.boxing(RealField, dim, dim)
+ // Nd4j specialized field.
+ val nd4jField = Nd4jArrayField.real(dim, dim)
measureAndPrint("Automatic field addition") {
autoField {
@@ -43,6 +49,13 @@ fun main() {
}
}
+ measureAndPrint("Nd4j specialized addition") {
+ nd4jField {
+ var res = one
+ repeat(n) { res += 1.0 as Number }
+ }
+ }
+
measureAndPrint("Lazy addition") {
val res = specializedField.one.mapAsync(GlobalScope) {
var c = 0.0
diff --git a/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstAlgebra.kt b/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstAlgebra.kt
index 9c58043dd..0e23b24a9 100644
--- a/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstAlgebra.kt
+++ b/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstAlgebra.kt
@@ -6,14 +6,13 @@ import kscience.kmath.operations.*
* [Algebra] over [MST] nodes.
*/
public object MstAlgebra : NumericAlgebra {
- override fun number(value: Number): MST.Numeric = MST.Numeric(value)
+ public override fun number(value: Number): MST.Numeric = MST.Numeric(value)
+ public override fun symbol(value: String): MST.Symbolic = MST.Symbolic(value)
- override fun symbol(value: String): MST.Symbolic = MST.Symbolic(value)
-
- override fun unaryOperation(operation: String, arg: MST): MST.Unary =
+ public override fun unaryOperation(operation: String, arg: MST): MST.Unary =
MST.Unary(operation, arg)
- override fun binaryOperation(operation: String, left: MST, right: MST): MST.Binary =
+ public override fun binaryOperation(operation: String, left: MST, right: MST): MST.Binary =
MST.Binary(operation, left, right)
}
@@ -33,7 +32,8 @@ public object MstSpace : Space, NumericAlgebra {
public override fun binaryOperation(operation: String, left: MST, right: MST): MST.Binary =
MstAlgebra.binaryOperation(operation, left, right)
- public override fun unaryOperation(operation: String, arg: MST): MST.Unary = MstAlgebra.unaryOperation(operation, arg)
+ public override fun unaryOperation(operation: String, arg: MST): MST.Unary =
+ MstAlgebra.unaryOperation(operation, arg)
}
/**
diff --git a/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstExpression.kt b/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstExpression.kt
index 5ca75e993..f68e3f5f8 100644
--- a/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstExpression.kt
+++ b/kmath-ast/src/commonMain/kotlin/kscience/kmath/ast/MstExpression.kt
@@ -13,7 +13,7 @@ import kotlin.contracts.contract
* @property mst the [MST] node.
* @author Alexander Nozik
*/
-public class MstExpression(public val algebra: Algebra, public val mst: MST) : Expression {
+public class MstExpression>(public val algebra: A, public val mst: MST) : Expression {
private inner class InnerAlgebra(val arguments: Map) : NumericAlgebra {
override fun symbol(value: String): T = arguments[StringSymbol(value)] ?: algebra.symbol(value)
override fun unaryOperation(operation: String, arg: T): T = algebra.unaryOperation(operation, arg)
@@ -21,8 +21,9 @@ public class MstExpression(public val algebra: Algebra, public val mst: MS
override fun binaryOperation(operation: String, left: T, right: T): T =
algebra.binaryOperation(operation, left, right)
- override fun number(value: Number): T = if (algebra is NumericAlgebra)
- algebra.number(value)
+ @Suppress("UNCHECKED_CAST")
+ override fun number(value: Number): T = if (algebra is NumericAlgebra<*>)
+ (algebra as NumericAlgebra).number(value)
else
error("Numeric nodes are not supported by $this")
}
@@ -38,14 +39,14 @@ public class MstExpression(public val algebra: Algebra, public val mst: MS
public inline fun , E : Algebra> A.mst(
mstAlgebra: E,
block: E.() -> MST,
-): MstExpression = MstExpression(this, mstAlgebra.block())
+): MstExpression = MstExpression(this, mstAlgebra.block())
/**
* Builds [MstExpression] over [Space].
*
* @author Alexander Nozik
*/
-public inline fun Space.mstInSpace(block: MstSpace.() -> MST): MstExpression {
+public inline fun > A.mstInSpace(block: MstSpace.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return MstExpression(this, MstSpace.block())
}
@@ -55,7 +56,7 @@ public inline fun Space.mstInSpace(block: MstSpace.() -> MS
*
* @author Alexander Nozik
*/
-public inline fun Ring.mstInRing(block: MstRing.() -> MST): MstExpression {
+public inline fun > A.mstInRing(block: MstRing.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return MstExpression(this, MstRing.block())
}
@@ -65,7 +66,7 @@ public inline fun Ring.mstInRing(block: MstRing.() -> MST):
*
* @author Alexander Nozik
*/
-public inline fun Field.mstInField(block: MstField.() -> MST): MstExpression {
+public inline fun > A.mstInField(block: MstField.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return MstExpression(this, MstField.block())
}
@@ -75,7 +76,7 @@ public inline fun Field.mstInField(block: MstField.() -> MS
*
* @author Iaroslav Postovalov
*/
-public inline fun Field.mstInExtendedField(block: MstExtendedField.() -> MST): MstExpression {
+public inline fun > A.mstInExtendedField(block: MstExtendedField.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return MstExpression(this, MstExtendedField.block())
}
@@ -85,7 +86,7 @@ public inline fun Field.mstInExtendedField(block: MstExtend
*
* @author Alexander Nozik
*/
-public inline fun > FunctionalExpressionSpace.mstInSpace(block: MstSpace.() -> MST): MstExpression {
+public inline fun > FunctionalExpressionSpace.mstInSpace(block: MstSpace.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return algebra.mstInSpace(block)
}
@@ -95,7 +96,7 @@ public inline fun > FunctionalExpressionSpace> FunctionalExpressionRing.mstInRing(block: MstRing.() -> MST): MstExpression {
+public inline fun > FunctionalExpressionRing.mstInRing(block: MstRing.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return algebra.mstInRing(block)
}
@@ -105,7 +106,7 @@ public inline fun > FunctionalExpressionRing.
*
* @author Alexander Nozik
*/
-public inline fun > FunctionalExpressionField.mstInField(block: MstField.() -> MST): MstExpression {
+public inline fun > FunctionalExpressionField.mstInField(block: MstField.() -> MST): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return algebra.mstInField(block)
}
@@ -117,7 +118,7 @@ public inline fun > FunctionalExpressionField> FunctionalExpressionExtendedField.mstInExtendedField(
block: MstExtendedField.() -> MST,
-): MstExpression {
+): MstExpression {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
return algebra.mstInExtendedField(block)
}
diff --git a/kmath-ast/src/jvmMain/kotlin/kscience/kmath/asm/asm.kt b/kmath-ast/src/jvmMain/kotlin/kscience/kmath/asm/asm.kt
index 2b6fa6247..9ccfa464c 100644
--- a/kmath-ast/src/jvmMain/kotlin/kscience/kmath/asm/asm.kt
+++ b/kmath-ast/src/jvmMain/kotlin/kscience/kmath/asm/asm.kt
@@ -69,4 +69,5 @@ public inline fun Algebra.expression(mst: MST): Expression<
*
* @author Alexander Nozik.
*/
-public inline fun MstExpression.compile(): Expression = mst.compileWith(T::class.java, algebra)
+public inline fun MstExpression>.compile(): Expression =
+ mst.compileWith(T::class.java, algebra)
diff --git a/kmath-commons/src/main/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpression.kt b/kmath-commons/src/main/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpression.kt
index c593f5103..345babe8b 100644
--- a/kmath-commons/src/main/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpression.kt
+++ b/kmath-commons/src/main/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpression.kt
@@ -12,16 +12,22 @@ import org.apache.commons.math3.analysis.differentiation.DerivativeStructure
*/
public class DerivativeStructureField(
public val order: Int,
- private val bindings: Map
+ bindings: Map,
) : ExtendedField, ExpressionAlgebra {
- public override val zero: DerivativeStructure by lazy { DerivativeStructure(bindings.size, order) }
- public override val one: DerivativeStructure by lazy { DerivativeStructure(bindings.size, order, 1.0) }
+ public val numberOfVariables: Int = bindings.size
+
+ public override val zero: DerivativeStructure by lazy { DerivativeStructure(numberOfVariables, order) }
+ public override val one: DerivativeStructure by lazy { DerivativeStructure(numberOfVariables, order, 1.0) }
/**
* A class that implements both [DerivativeStructure] and a [Symbol]
*/
- public inner class DerivativeStructureSymbol(symbol: Symbol, value: Double) :
- DerivativeStructure(bindings.size, order, bindings.keys.indexOf(symbol), value), Symbol {
+ public inner class DerivativeStructureSymbol(
+ size: Int,
+ index: Int,
+ symbol: Symbol,
+ value: Double,
+ ) : DerivativeStructure(size, order, index, value), Symbol {
override val identity: String = symbol.identity
override fun toString(): String = identity
override fun equals(other: Any?): Boolean = this.identity == (other as? Symbol)?.identity
@@ -31,27 +37,26 @@ public class DerivativeStructureField(
/**
* Identity-based symbol bindings map
*/
- private val variables: Map = bindings.entries.associate { (key, value) ->
- key.identity to DerivativeStructureSymbol(key, value)
- }
+ private val variables: Map = bindings.entries.mapIndexed { index, (key, value) ->
+ key.identity to DerivativeStructureSymbol(numberOfVariables, index, key, value)
+ }.toMap()
- override fun const(value: Double): DerivativeStructure = DerivativeStructure(bindings.size, order, value)
+ override fun const(value: Double): DerivativeStructure = DerivativeStructure(numberOfVariables, order, value)
public override fun bindOrNull(symbol: Symbol): DerivativeStructureSymbol? = variables[symbol.identity]
public fun bind(symbol: Symbol): DerivativeStructureSymbol = variables.getValue(symbol.identity)
- //public fun Number.const(): DerivativeStructure = const(toDouble())
+ override fun symbol(value: String): DerivativeStructureSymbol = bind(StringSymbol(value))
- public fun DerivativeStructure.derivative(parameter: Symbol, order: Int = 1): Double {
- return derivative(mapOf(parameter to order))
+ public fun DerivativeStructure.derivative(symbols: List): Double {
+ require(symbols.size <= order) { "The order of derivative ${symbols.size} exceeds computed order $order" }
+ val ordersCount = symbols.map { it.identity }.groupBy { it }.mapValues { it.value.size }
+ return getPartialDerivative(*variables.keys.map { ordersCount[it] ?: 0 }.toIntArray())
}
- public fun DerivativeStructure.derivative(orders: Map): Double {
- return getPartialDerivative(*bindings.keys.map { orders[it] ?: 0 }.toIntArray())
- }
+ public fun DerivativeStructure.derivative(vararg symbols: Symbol): Double = derivative(symbols.toList())
- public fun DerivativeStructure.derivative(vararg orders: Pair): Double = derivative(mapOf(*orders))
public override fun add(a: DerivativeStructure, b: DerivativeStructure): DerivativeStructure = a.add(b)
public override fun multiply(a: DerivativeStructure, k: Number): DerivativeStructure = when (k) {
@@ -90,26 +95,27 @@ public class DerivativeStructureField(
public override operator fun Number.plus(b: DerivativeStructure): DerivativeStructure = b + this
public override operator fun Number.minus(b: DerivativeStructure): DerivativeStructure = b - this
- public companion object : AutoDiffProcessor {
- override fun process(function: DerivativeStructureField.() -> DerivativeStructure): DifferentiableExpression {
- return DerivativeStructureExpression(function)
- }
+ public companion object :
+ AutoDiffProcessor> {
+ public override fun process(function: DerivativeStructureField.() -> DerivativeStructure): DifferentiableExpression> =
+ DerivativeStructureExpression(function)
}
}
+
/**
* A constructs that creates a derivative structure with required order on-demand
*/
public class DerivativeStructureExpression(
public val function: DerivativeStructureField.() -> DerivativeStructure,
-) : DifferentiableExpression {
+) : DifferentiableExpression> {
public override operator fun invoke(arguments: Map): Double =
DerivativeStructureField(0, arguments).function().value
/**
* Get the derivative expression with given orders
*/
- public override fun derivativeOrNull(orders: Map): Expression = Expression { arguments ->
- with(DerivativeStructureField(orders.values.maxOrNull() ?: 0, arguments)) { function().derivative(orders) }
+ public override fun derivativeOrNull(symbols: List): Expression = Expression { arguments ->
+ with(DerivativeStructureField(symbols.size, arguments)) { function().derivative(symbols) }
}
}
diff --git a/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/CMOptimizationProblem.kt b/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/CMOptimizationProblem.kt
index 13f9af7bb..d6f79529a 100644
--- a/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/CMOptimizationProblem.kt
+++ b/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/CMOptimizationProblem.kt
@@ -19,9 +19,8 @@ import kotlin.reflect.KClass
public operator fun PointValuePair.component1(): DoubleArray = point
public operator fun PointValuePair.component2(): Double = value
-public class CMOptimizationProblem(
- override val symbols: List,
-) : OptimizationProblem, SymbolIndexer, OptimizationFeature {
+public class CMOptimizationProblem(override val symbols: List, ) :
+ OptimizationProblem, SymbolIndexer, OptimizationFeature {
private val optimizationData: HashMap, OptimizationData> = HashMap()
private var optimizatorBuilder: (() -> MultivariateOptimizer)? = null
public var convergenceChecker: ConvergenceChecker = SimpleValueChecker(DEFAULT_RELATIVE_TOLERANCE,
@@ -49,7 +48,7 @@ public class CMOptimizationProblem(
addOptimizationData(objectiveFunction)
}
- public override fun diffExpression(expression: DifferentiableExpression): Unit {
+ public override fun diffExpression(expression: DifferentiableExpression>) {
expression(expression)
val gradientFunction = ObjectiveFunctionGradient {
val args = it.toMap()
diff --git a/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/cmFit.kt b/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/cmFit.kt
index 42475db6c..b8e8bfd4b 100644
--- a/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/cmFit.kt
+++ b/kmath-commons/src/main/kotlin/kscience/kmath/commons/optimization/cmFit.kt
@@ -12,7 +12,6 @@ import kscience.kmath.structures.asBuffer
import org.apache.commons.math3.analysis.differentiation.DerivativeStructure
import org.apache.commons.math3.optim.nonlinear.scalar.GoalType
-
/**
* Generate a chi squared expression from given x-y-sigma data and inline model. Provides automatic differentiation
*/
@@ -21,7 +20,7 @@ public fun Fitting.chiSquared(
y: Buffer,
yErr: Buffer,
model: DerivativeStructureField.(x: DerivativeStructure) -> DerivativeStructure,
-): DifferentiableExpression = chiSquared(DerivativeStructureField, x, y, yErr, model)
+): DifferentiableExpression> = chiSquared(DerivativeStructureField, x, y, yErr, model)
/**
* Generate a chi squared expression from given x-y-sigma data and inline model. Provides automatic differentiation
@@ -31,7 +30,7 @@ public fun Fitting.chiSquared(
y: Iterable,
yErr: Iterable,
model: DerivativeStructureField.(x: DerivativeStructure) -> DerivativeStructure,
-): DifferentiableExpression = chiSquared(
+): DifferentiableExpression> = chiSquared(
DerivativeStructureField,
x.toList().asBuffer(),
y.toList().asBuffer(),
@@ -39,7 +38,6 @@ public fun Fitting.chiSquared(
model
)
-
/**
* Optimize expression without derivatives
*/
@@ -48,16 +46,15 @@ public fun Expression.optimize(
configuration: CMOptimizationProblem.() -> Unit,
): OptimizationResult = optimizeWith(CMOptimizationProblem, symbols = symbols, configuration)
-
/**
* Optimize differentiable expression
*/
-public fun DifferentiableExpression.optimize(
+public fun DifferentiableExpression>.optimize(
vararg symbols: Symbol,
configuration: CMOptimizationProblem.() -> Unit,
): OptimizationResult = optimizeWith(CMOptimizationProblem, symbols = symbols, configuration)
-public fun DifferentiableExpression.minimize(
+public fun DifferentiableExpression>.minimize(
vararg startPoint: Pair,
configuration: CMOptimizationProblem.() -> Unit = {},
): OptimizationResult {
diff --git a/kmath-commons/src/test/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpressionTest.kt b/kmath-commons/src/test/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpressionTest.kt
index 8886e123f..7511a38ed 100644
--- a/kmath-commons/src/test/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpressionTest.kt
+++ b/kmath-commons/src/test/kotlin/kscience/kmath/commons/expressions/DerivativeStructureExpressionTest.kt
@@ -5,14 +5,15 @@ import kotlin.contracts.InvocationKind
import kotlin.contracts.contract
import kotlin.test.Test
import kotlin.test.assertEquals
+import kotlin.test.assertFails
-internal inline fun diff(
+internal inline fun diff(
order: Int,
vararg parameters: Pair,
- block: DerivativeStructureField.() -> R,
-): R {
+ block: DerivativeStructureField.() -> Unit,
+): Unit {
contract { callsInPlace(block, InvocationKind.EXACTLY_ONCE) }
- return DerivativeStructureField(order, mapOf(*parameters)).run(block)
+ DerivativeStructureField(order, mapOf(*parameters)).run(block)
}
internal class AutoDiffTest {
@@ -21,13 +22,16 @@ internal class AutoDiffTest {
@Test
fun derivativeStructureFieldTest() {
- val res: Double = diff(3, x to 1.0, y to 1.0) {
+ diff(2, x to 1.0, y to 1.0) {
val x = bind(x)//by binding()
val y = symbol("y")
- val z = x * (-sin(x * y) + y)
- z.derivative(x)
+ val z = x * (-sin(x * y) + y) + 2.0
+ println(z.derivative(x))
+ println(z.derivative(y,x))
+ assertEquals(z.derivative(x, y), z.derivative(y, x))
+ //check that improper order cause failure
+ assertFails { z.derivative(x,x,y) }
}
- println(res)
}
@Test
@@ -40,5 +44,7 @@ internal class AutoDiffTest {
assertEquals(10.0, f(x to 1.0, y to 2.0))
assertEquals(6.0, f.derivative(x)(x to 1.0, y to 2.0))
+ assertEquals(2.0, f.derivative(x, x)(x to 1.234, y to -2.0))
+ assertEquals(2.0, f.derivative(x, y)(x to 1.0, y to 2.0))
}
}
diff --git a/kmath-commons/src/test/kotlin/kscience/kmath/commons/optimization/OptimizeTest.kt b/kmath-commons/src/test/kotlin/kscience/kmath/commons/optimization/OptimizeTest.kt
index 4384a5124..3290c8f32 100644
--- a/kmath-commons/src/test/kotlin/kscience/kmath/commons/optimization/OptimizeTest.kt
+++ b/kmath-commons/src/test/kotlin/kscience/kmath/commons/optimization/OptimizeTest.kt
@@ -6,7 +6,6 @@ import kscience.kmath.stat.Distribution
import kscience.kmath.stat.Fitting
import kscience.kmath.stat.RandomGenerator
import kscience.kmath.stat.normal
-import kscience.kmath.structures.asBuffer
import org.junit.jupiter.api.Test
import kotlin.math.pow
@@ -48,14 +47,17 @@ internal class OptimizeTest {
val sigma = 1.0
val generator = Distribution.normal(0.0, sigma)
val chain = generator.sample(RandomGenerator.default(112667))
- val x = (1..100).map { it.toDouble() }
- val y = x.map { it ->
+ val x = (1..100).map(Int::toDouble)
+
+ val y = x.map {
it.pow(2) + it + 1 + chain.nextDouble()
}
- val yErr = x.map { sigma }
- val chi2 = Fitting.chiSquared(x.asBuffer(), y.asBuffer(), yErr.asBuffer()) { x ->
+
+ val yErr = List(x.size) { sigma }
+
+ val chi2 = Fitting.chiSquared(x, y, yErr) { x1 ->
val cWithDefault = bindOrNull(c) ?: one
- bind(a) * x.pow(2) + bind(b) * x + cWithDefault
+ bind(a) * x1.pow(2) + bind(b) * x1 + cWithDefault
}
val result = chi2.minimize(a to 1.5, b to 0.9, c to 1.0)
diff --git a/kmath-core/README.md b/kmath-core/README.md
index 5501b1d7a..42a513a10 100644
--- a/kmath-core/README.md
+++ b/kmath-core/README.md
@@ -12,7 +12,7 @@ The core features of KMath:
> #### Artifact:
>
-> This module artifact: `kscience.kmath:kmath-core:0.2.0-dev-2`.
+> This module artifact: `kscience.kmath:kmath-core:0.2.0-dev-3`.
>
> 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-2'
+> implementation 'kscience.kmath:kmath-core:0.2.0-dev-3'
> }
> ```
> **Gradle Kotlin DSL:**
@@ -44,6 +44,6 @@ The core features of KMath:
> }
>
> dependencies {
-> implementation("kscience.kmath:kmath-core:0.2.0-dev-2")
+> implementation("kscience.kmath:kmath-core:0.2.0-dev-3")
> }
> ```
diff --git a/kmath-core/build.gradle.kts b/kmath-core/build.gradle.kts
index b0849eca5..7f889d9b4 100644
--- a/kmath-core/build.gradle.kts
+++ b/kmath-core/build.gradle.kts
@@ -1,3 +1,5 @@
+import ru.mipt.npm.gradle.Maturity
+
plugins {
id("ru.mipt.npm.mpp")
id("ru.mipt.npm.native")
@@ -11,36 +13,42 @@ kotlin.sourceSets.commonMain {
readme {
description = "Core classes, algebra definitions, basic linear algebra"
- maturity = ru.mipt.npm.gradle.Maturity.DEVELOPMENT
+ maturity = Maturity.DEVELOPMENT
propertyByTemplate("artifact", rootProject.file("docs/templates/ARTIFACT-TEMPLATE.md"))
+
feature(
id = "algebras",
description = "Algebraic structures: contexts and elements",
ref = "src/commonMain/kotlin/kscience/kmath/operations/Algebra.kt"
)
+
feature(
id = "nd",
description = "Many-dimensional structures",
ref = "src/commonMain/kotlin/kscience/kmath/structures/NDStructure.kt"
)
+
feature(
id = "buffers",
description = "One-dimensional structure",
ref = "src/commonMain/kotlin/kscience/kmath/structures/Buffers.kt"
)
+
feature(
id = "expressions",
description = "Functional Expressions",
ref = "src/commonMain/kotlin/kscience/kmath/expressions"
)
+
feature(
id = "domains",
description = "Domains",
ref = "src/commonMain/kotlin/kscience/kmath/domains"
)
+
feature(
id = "autodif",
description = "Automatic differentiation",
ref = "src/commonMain/kotlin/kscience/kmath/expressions/SimpleAutoDiff.kt"
)
-}
\ No newline at end of file
+}
diff --git a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/DifferentiableExpression.kt b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/DifferentiableExpression.kt
index 4fe73f283..abce9c4ec 100644
--- a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/DifferentiableExpression.kt
+++ b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/DifferentiableExpression.kt
@@ -1,32 +1,41 @@
package kscience.kmath.expressions
/**
- * An expression that provides derivatives
+ * Represents expression which structure can be differentiated.
+ *
+ * @param T the type this expression takes as argument and returns.
+ * @param R the type of expression this expression can be differentiated to.
*/
-public interface DifferentiableExpression : Expression{
- public fun derivativeOrNull(orders: Map): Expression?
+public interface DifferentiableExpression> : Expression {
+ /**
+ * Differentiates this expression by ordered collection of [symbols].
+ *
+ * @param symbols the symbols.
+ * @return the derivative or `null`.
+ */
+ public fun derivativeOrNull(symbols: List): R?
}
-public fun DifferentiableExpression.derivative(orders: Map): Expression =
- derivativeOrNull(orders) ?: error("Derivative with orders $orders not provided")
+public fun > DifferentiableExpression.derivative(symbols: List): R =
+ derivativeOrNull(symbols) ?: error("Derivative by symbols $symbols not provided")
-public fun DifferentiableExpression.derivative(vararg orders: Pair): Expression =
- derivative(mapOf(*orders))
+public fun > DifferentiableExpression.derivative(vararg symbols: Symbol): R =
+ derivative(symbols.toList())
-public fun DifferentiableExpression.derivative(symbol: Symbol): Expression = derivative(symbol to 1)
-
-public fun DifferentiableExpression.derivative(name: String): Expression =
- derivative(StringSymbol(name) to 1)
+public fun > DifferentiableExpression.derivative(name: String): R =
+ derivative(StringSymbol(name))
/**
* A [DifferentiableExpression] that defines only first derivatives
*/
-public abstract class FirstDerivativeExpression : DifferentiableExpression {
+public abstract class FirstDerivativeExpression> : DifferentiableExpression {
+ /**
+ * Returns first derivative of this expression by given [symbol].
+ */
+ public abstract fun derivativeOrNull(symbol: Symbol): R?
- public abstract fun derivativeOrNull(symbol: Symbol): Expression?
-
- public override fun derivativeOrNull(orders: Map): Expression? {
- val dSymbol = orders.entries.singleOrNull { it.value == 1 }?.key ?: return null
+ public final override fun derivativeOrNull(symbols: List): R? {
+ val dSymbol = symbols.firstOrNull() ?: return null
return derivativeOrNull(dSymbol)
}
}
@@ -34,6 +43,6 @@ public abstract class FirstDerivativeExpression : DifferentiableExpression
/**
* A factory that converts an expression in autodiff variables to a [DifferentiableExpression]
*/
-public interface AutoDiffProcessor> {
- public fun process(function: A.() -> I): DifferentiableExpression
-}
\ No newline at end of file
+public fun interface AutoDiffProcessor, out R : Expression> {
+ public fun process(function: A.() -> I): DifferentiableExpression
+}
diff --git a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/Expression.kt b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/Expression.kt
index ab9ff0e72..98940e767 100644
--- a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/Expression.kt
+++ b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/Expression.kt
@@ -22,7 +22,9 @@ public inline class StringSymbol(override val identity: String) : Symbol {
}
/**
- * An elementary function that could be invoked on a map of arguments
+ * An elementary function that could be invoked on a map of arguments.
+ *
+ * @param T the type this expression takes as argument and returns.
*/
public fun interface Expression {
/**
@@ -35,20 +37,27 @@ public fun interface Expression {
}
/**
- * Invoke an expression without parameters
+ * Calls this expression without providing any arguments.
+ *
+ * @return a value.
*/
public operator fun Expression.invoke(): T = invoke(emptyMap())
-//This method exists to avoid resolution ambiguity of vararg methods
/**
* Calls this expression from arguments.
*
- * @param pairs the pair of arguments' names to values.
- * @return the value.
+ * @param pairs the pairs of arguments to values.
+ * @return a value.
*/
@JvmName("callBySymbol")
public operator fun Expression.invoke(vararg pairs: Pair): T = invoke(mapOf(*pairs))
+/**
+ * Calls this expression from arguments.
+ *
+ * @param pairs the pairs of arguments' names to values.
+ * @return a value.
+ */
@JvmName("callByString")
public operator fun Expression.invoke(vararg pairs: Pair): T =
invoke(mapOf(*pairs).mapKeys { StringSymbol(it.key) })
@@ -61,7 +70,6 @@ public operator fun Expression.invoke(vararg pairs: Pair): T =
* @param E type of the actual expression state
*/
public interface ExpressionAlgebra : Algebra {
-
/**
* Bind a given [Symbol] to this context variable and produce context-specific object. Return null if symbol could not be bound in current context.
*/
@@ -87,7 +95,7 @@ public fun ExpressionAlgebra.bind(symbol: Symbol): E =
/**
* A delegate to create a symbol with a string identity in this scope
*/
-public val symbol: ReadOnlyProperty = ReadOnlyProperty { thisRef, property ->
+public val symbol: ReadOnlyProperty = ReadOnlyProperty { _, property ->
StringSymbol(property.name)
}
@@ -96,4 +104,4 @@ public val symbol: ReadOnlyProperty = ReadOnlyProperty {
*/
public fun ExpressionAlgebra.binding(): ReadOnlyProperty = ReadOnlyProperty { _, property ->
bind(StringSymbol(property.name)) ?: error("A variable with name ${property.name} does not exist")
-}
\ No newline at end of file
+}
diff --git a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/SimpleAutoDiff.kt b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/SimpleAutoDiff.kt
index 5a9642690..e8a894d23 100644
--- a/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/SimpleAutoDiff.kt
+++ b/kmath-core/src/commonMain/kotlin/kscience/kmath/expressions/SimpleAutoDiff.kt
@@ -68,7 +68,7 @@ public fun > F.simpleAutoDiff(
): DerivationResult {
contract { callsInPlace(body, InvocationKind.EXACTLY_ONCE) }
- return SimpleAutoDiffField(this, bindings).derivate(body)
+ return SimpleAutoDiffField(this, bindings).differentiate(body)
}
public fun > F.simpleAutoDiff(
@@ -83,12 +83,21 @@ public open class SimpleAutoDiffField>(
public val context: F,
bindings: Map,
) : Field>, ExpressionAlgebra> {
+ public override val zero: AutoDiffValue
+ get() = const(context.zero)
+
+ public override val one: AutoDiffValue
+ get() = const(context.one)
// this stack contains pairs of blocks and values to apply them to
private var stack: Array = arrayOfNulls(8)
private var sp: Int = 0
private val derivatives: MutableMap, T> = hashMapOf()
+ private val bindings: Map> = bindings.entries.associate {
+ it.key.identity to AutoDiffVariableWithDerivative(it.key.identity, it.value, context.zero)
+ }
+
/**
* Differentiable variable with value and derivative of differentiation ([simpleAutoDiff]) result
* with respect to this variable.
@@ -106,11 +115,7 @@ public open class SimpleAutoDiffField>(
override fun hashCode(): Int = identity.hashCode()
}
- private val bindings: Map> = bindings.entries.associate {
- it.key.identity to AutoDiffVariableWithDerivative(it.key.identity, it.value, context.zero)
- }
-
- override fun bindOrNull(symbol: Symbol): AutoDiffValue? = bindings[symbol.identity]
+ public override fun bindOrNull(symbol: Symbol): AutoDiffValue? = bindings[symbol.identity]
private fun getDerivative(variable: AutoDiffValue): T =
(variable as? AutoDiffVariableWithDerivative)?.d ?: derivatives[variable] ?: context.zero
@@ -119,7 +124,6 @@ public open class SimpleAutoDiffField>(
if (variable is AutoDiffVariableWithDerivative) variable.d = value else derivatives[variable] = value
}
-
@Suppress("UNCHECKED_CAST")
private fun runBackwardPass() {
while (sp > 0) {
@@ -129,9 +133,6 @@ public open class SimpleAutoDiffField>(
}
}
- override val zero: AutoDiffValue get() = const(context.zero)
- override val one: AutoDiffValue get() = const(context.one)
-
override fun const(value: T): AutoDiffValue = AutoDiffValue(value)
/**
@@ -165,7 +166,7 @@ public open class SimpleAutoDiffField>(
}
- internal fun derivate(function: SimpleAutoDiffField.() -> AutoDiffValue): DerivationResult {
+ internal fun differentiate(function: SimpleAutoDiffField.() -> AutoDiffValue): DerivationResult {
val result = function()
result.d = context.one // computing derivative w.r.t result
runBackwardPass()
@@ -174,41 +175,41 @@ public open class SimpleAutoDiffField>(
// Overloads for Double constants
- override operator fun Number.plus(b: AutoDiffValue): AutoDiffValue =
+ public override operator fun Number.plus(b: AutoDiffValue): AutoDiffValue =
derive(const { this@plus.toDouble() * one + b.value }) { z ->
b.d += z.d
}
- override operator fun AutoDiffValue.plus(b: Number): AutoDiffValue = b.plus(this)
+ public override operator fun AutoDiffValue.plus(b: Number): AutoDiffValue = b.plus(this)
- override operator fun Number.minus(b: AutoDiffValue): AutoDiffValue =
+ public override operator fun Number.minus(b: AutoDiffValue): AutoDiffValue =
derive(const { this@minus.toDouble() * one - b.value }) { z -> b.d -= z.d }
- override operator fun AutoDiffValue.minus(b: Number): AutoDiffValue =
+ public override operator fun AutoDiffValue.minus(b: Number): AutoDiffValue =
derive(const { this@minus.value - one * b.toDouble() }) { z -> this@minus.d += z.d }
// Basic math (+, -, *, /)
- override fun add(a: AutoDiffValue, b: AutoDiffValue): AutoDiffValue =
+ public override fun add(a: AutoDiffValue, b: AutoDiffValue): AutoDiffValue =
derive(const { a.value + b.value }) { z ->
a.d += z.d
b.d += z.d
}
- override fun multiply(a: AutoDiffValue, b: AutoDiffValue