add native statistics impl #62

multik-core/build.gradle.kts

@@ -70,13 +70,13 @@ kotlin {
                 implementation("org.apache.commons:commons-csv:$common_csv_version")
             }
         }
-        val nativeCommonMain by creating {
+        val nativeMain by creating {
             dependsOn(commonMain)
         }
         names.forEach { n ->
             if (n.contains("X64Main") || n.contains("Arm64Main")){
                 this@sourceSets.getByName(n).apply{
-                    dependsOn(nativeCommonMain)
+                    dependsOn(nativeMain)
                 }
             }
         }

multik-core/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/api/Engine.kt

@@ -6,6 +6,7 @@ package org.jetbrains.kotlinx.multik.api
 
 import org.jetbrains.kotlinx.multik.api.linalg.LinAlg
 import org.jetbrains.kotlinx.multik.api.math.Math
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 
 public sealed class EngineType(public val name: String)
 

multik-core/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/api/Multik.kt

@@ -11,6 +11,7 @@ import org.jetbrains.kotlinx.multik.api.Multik.math
 import org.jetbrains.kotlinx.multik.api.Multik.stat
 import org.jetbrains.kotlinx.multik.api.linalg.LinAlg
 import org.jetbrains.kotlinx.multik.api.math.Math
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 
 /**
  * Abbreviated name for [Multik].

multik-core/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/api/Statistics.kt → multik-core/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/api/stat/Statistics.kt

@@ -1,8 +1,8 @@
 /*
- * Copyright 2020-2021 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
-package org.jetbrains.kotlinx.multik.api
+package org.jetbrains.kotlinx.multik.api.stat
 
 import org.jetbrains.kotlinx.multik.ndarray.complex.ComplexDouble
 import org.jetbrains.kotlinx.multik.ndarray.complex.ComplexFloat

multik-default/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/default/DefaultEngine.kt

@@ -4,11 +4,15 @@
 
 package org.jetbrains.kotlinx.multik.default
 
-import org.jetbrains.kotlinx.multik.api.*
+import org.jetbrains.kotlinx.multik.api.DefaultEngineType
+import org.jetbrains.kotlinx.multik.api.Engine
+import org.jetbrains.kotlinx.multik.api.EngineType
 import org.jetbrains.kotlinx.multik.api.linalg.LinAlg
 import org.jetbrains.kotlinx.multik.api.math.Math
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 import org.jetbrains.kotlinx.multik.default.linalg.DefaultLinAlg
 import org.jetbrains.kotlinx.multik.default.math.DefaultMath
+import org.jetbrains.kotlinx.multik.default.stat.DefaultStatistics
 
 public class DefaultEngine : Engine() {
     override val name: String

multik-default/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/default/DefaultStatistics.kt → multik-default/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/default/stat/DefaultStatistics.kt

@@ -2,30 +2,26 @@
  * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
-package org.jetbrains.kotlinx.multik.default
+package org.jetbrains.kotlinx.multik.default.stat
 
-import org.jetbrains.kotlinx.multik.api.Statistics
-import org.jetbrains.kotlinx.multik.kotlin.KEStatistics
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 import org.jetbrains.kotlinx.multik.ndarray.data.*
 
-public object DefaultStatistics : Statistics {
+public expect object DefaultStatistics : Statistics {
 
-    override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double? = KEStatistics.median(a)
+    override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double?
 
-    override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double =
-        KEStatistics.average(a, weights)
+    override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double
 
-    override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double = KEStatistics.mean(a)
+    override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double
 
-    override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O> =
-        KEStatistics.mean(a, axis)
+    override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O>
 
-    override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1> = mean(a, axis)
+    override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1>
 
-    override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2> = mean(a, axis)
+    override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2>
 
-    override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3> = mean(a, axis)
-
-    override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4> = mean(a, axis)
+    override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3>
 
+    override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4>
 }

multik-default/src/iosMain/kotlin/org.jetbrains.kotlinx.multik.default/stat/DefaultStatistics.kt

@@ -0,0 +1,30 @@
+/*
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ */
+
+package org.jetbrains.kotlinx.multik.default.stat
+
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
+import org.jetbrains.kotlinx.multik.ndarray.data.*
+
+public actual object DefaultStatistics : Statistics {
+
+    actual override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double? = KEStatistics.median(a)
+
+    actual override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double =
+        KEStatistics.average(a, weights)
+
+    actual override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double = KEStatistics.mean(a)
+
+    actual override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O> =
+        KEStatistics.mean(a, axis)
+
+    actual override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1> = mean(a, axis)
+
+    actual override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2> = mean(a, axis)
+
+    actual override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3> = mean(a, axis)
+
+    actual override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4> = mean(a, axis)
+}

multik-default/src/jsMain/kotlin/org.jetbrains.kotlinx.multik.default/stat/DefaultStatistics.kt

@@ -0,0 +1,30 @@
+/*
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ */
+
+package org.jetbrains.kotlinx.multik.default.stat
+
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
+import org.jetbrains.kotlinx.multik.ndarray.data.*
+
+public actual object DefaultStatistics : Statistics {
+
+    actual override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double? = KEStatistics.median(a)
+
+    actual override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double =
+        KEStatistics.average(a, weights)
+
+    actual override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double = KEStatistics.mean(a)
+
+    actual override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O> =
+        KEStatistics.mean(a, axis)
+
+    actual override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1> = mean(a, axis)
+
+    actual override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2> = mean(a, axis)
+
+    actual override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3> = mean(a, axis)
+
+    actual override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4> = mean(a, axis)
+}

multik-default/src/jvmMain/kotlin/org/jetbrains/kotlinx/multik/default/stat/DefaultStatistics.kt

@@ -0,0 +1,31 @@
+/*
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ */
+
+package org.jetbrains.kotlinx.multik.default.stat
+
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
+import org.jetbrains.kotlinx.multik.ndarray.data.*
+import org.jetbrains.kotlinx.multik.openblas.stat.NativeStatistics
+
+public actual object DefaultStatistics : Statistics {
+
+    actual override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double? = NativeStatistics.median(a)
+
+    actual override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double =
+        KEStatistics.average(a, weights)
+
+    actual override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double = KEStatistics.mean(a)
+
+    actual override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O> =
+        KEStatistics.mean(a, axis)
+
+    actual override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1> = mean(a, axis)
+
+    actual override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2> = mean(a, axis)
+
+    actual override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3> = mean(a, axis)
+
+    actual override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4> = mean(a, axis)
+}

multik-default/src/nativeMain/kotlin/org.jetbrains.kotlinx.multik.default/stat/DefaultStatistics.kt

@@ -0,0 +1,31 @@
+/*
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ */
+
+package org.jetbrains.kotlinx.multik.default.stat
+
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
+import org.jetbrains.kotlinx.multik.ndarray.data.*
+import org.jetbrains.kotlinx.multik.openblas.stat.NativeStatistics
+
+public actual object DefaultStatistics : Statistics {
+
+    actual override fun <T : Number, D : Dimension> median(a: MultiArray<T, D>): Double? = NativeStatistics.median(a)
+
+    actual override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double =
+        KEStatistics.average(a, weights)
+
+    actual override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double = KEStatistics.mean(a)
+
+    actual override fun <T : Number, D : Dimension, O : Dimension> mean(a: MultiArray<T, D>, axis: Int): NDArray<Double, O> =
+        KEStatistics.mean(a, axis)
+
+    actual override fun <T : Number> meanD2(a: MultiArray<T, D2>, axis: Int): NDArray<Double, D1> = mean(a, axis)
+
+    actual override fun <T : Number> meanD3(a: MultiArray<T, D3>, axis: Int): NDArray<Double, D2> = mean(a, axis)
+
+    actual override fun <T : Number> meanD4(a: MultiArray<T, D4>, axis: Int): NDArray<Double, D3> = mean(a, axis)
+
+    actual override fun <T : Number> meanDN(a: MultiArray<T, DN>, axis: Int): NDArray<Double, D4> = mean(a, axis)
+}

multik-kotlin/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/kotlin/KEEngine.kt

@@ -1,17 +1,18 @@
 /*
- * Copyright 2020-2021 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
 package org.jetbrains.kotlinx.multik.kotlin
 
 import org.jetbrains.kotlinx.multik.api.Engine
 import org.jetbrains.kotlinx.multik.api.EngineType
 import org.jetbrains.kotlinx.multik.api.KEEngineType
-import org.jetbrains.kotlinx.multik.api.Statistics
 import org.jetbrains.kotlinx.multik.api.linalg.LinAlg
 import org.jetbrains.kotlinx.multik.api.math.Math
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 import org.jetbrains.kotlinx.multik.kotlin.linalg.KELinAlg
 import org.jetbrains.kotlinx.multik.kotlin.math.KEMath
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
 
 
 public class KEEngine : Engine() {

multik-kotlin/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/kotlin/KEStatistics.kt → multik-kotlin/src/commonMain/kotlin/org/jetbrains/kotlinx/multik/kotlin/stat/KEStatistics.kt

@@ -1,11 +1,10 @@
 /*
- * Copyright 2020-2021 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
-package org.jetbrains.kotlinx.multik.kotlin
+package org.jetbrains.kotlinx.multik.kotlin.stat
 
-import org.jetbrains.kotlinx.multik.api.Statistics
-import org.jetbrains.kotlinx.multik.api.mk
+import org.jetbrains.kotlinx.multik.api.stat.Statistics
 import org.jetbrains.kotlinx.multik.kotlin.math.KEMath
 import org.jetbrains.kotlinx.multik.kotlin.math.remove
 import org.jetbrains.kotlinx.multik.ndarray.data.*
@@ -34,7 +33,7 @@ public object KEStatistics : Statistics {
 
     override fun <T : Number, D : Dimension> average(a: MultiArray<T, D>, weights: MultiArray<T, D>?): Double {
         if (weights == null) return mean(a)
-        return mk.math.sum(a * weights).toDouble() / mk.math.sum(weights).toDouble()
+        return KEMath.sum(a * weights).toDouble() / KEMath.sum(weights).toDouble()
     }
 
     override fun <T : Number, D : Dimension> mean(a: MultiArray<T, D>): Double {

multik-kotlin/src/commonTest/kotlin/org/jetbrains/kotlinx/multik_kotlin/statistics/KEStatisticsTest.kt

@@ -1,13 +1,13 @@
 /*
- * Copyright 2020-2021 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
 package org.jetbrains.kotlinx.multik_kotlin.statistics
 
 import org.jetbrains.kotlinx.multik.api.arange
 import org.jetbrains.kotlinx.multik.api.mk
 import org.jetbrains.kotlinx.multik.api.ndarray
-import org.jetbrains.kotlinx.multik.kotlin.KEStatistics
+import org.jetbrains.kotlinx.multik.kotlin.stat.KEStatistics
 import kotlin.test.Test
 import kotlin.test.assertEquals
 
@@ -16,7 +16,7 @@ class KEStatisticsTest {
     @Test
     fun test_median() {
         val a = mk.ndarray(mk[mk[10, 7, 4], mk[3, 2, 1]])
-        println(mk.stat.median(a))
+        assertEquals(3.5, mk.stat.median(a))
     }
 
     @Test

multik-openblas/build.gradle.kts

@@ -72,7 +72,7 @@ kotlin {
                     val cinteropDir = "${projectDir}/cinterop"
                     val headersDir = "${projectDir}/multik_jni/src/main/headers/"
                     val cppDir = "${projectDir}/multik_jni/src/main/cpp"
-                    headers("$headersDir/mk_math.h", "$headersDir/mk_linalg.h")
+                    headers("$headersDir/mk_math.h", "$headersDir/mk_linalg.h", "$headersDir/mk_stat.h")
                     defFile(project.file(("$cinteropDir/libmultik.def")))
 
                     when (konanTarget.family) {
@@ -87,6 +87,7 @@ kotlin {
                     extraOpts("-Xsource-compiler-option", "-I${buildDir}/cmake-build/openblas-install/include")
                     extraOpts("-Xcompile-source", "$cppDir/mk_math.cpp")
                     extraOpts("-Xcompile-source", "$cppDir/mk_linalg.cpp")
+                    extraOpts("-Xcompile-source", "$cppDir/mk_stat.cpp")
                 }
             }
         }

multik-openblas/multik_jni/CMakeLists.txt

@@ -114,14 +114,17 @@ include_directories("src/main/headers")
 set(SRC_FILES_FOR_SHARED
         src/main/cpp/jni_Linalg.cpp
         src/main/cpp/jni_JniMath.cpp
+        src/main/cpp/JniStat.cpp
         src/main/cpp/ComplexDouble.cpp
         src/main/cpp/ComplexFloat.cpp
         src/main/cpp/mk_math.cpp
         src/main/cpp/mk_linalg.cpp
+        src/main/cpp/mk_stat.cpp
         )
 set(SRC_FILES_FOR_STATIC
         src/main/cpp/mk_math.cpp
-        src/main/cpp/mk_linalg.cpp)
+        src/main/cpp/mk_linalg.cpp
+        src/main/cpp/mk_stat.cpp)
 ########################
 
 ### FIND GFORTRAN AND QUADMATH LIBRARIES ###

multik-openblas/multik_jni/src/main/cpp/JniStat.cpp

@@ -0,0 +1,23 @@
+/*
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ */
+
+#include "JniStat.h"
+#include "mk_stat.h"
+
+/*
+ * Class:     org_jetbrains_kotlinx_multik_openblas_stat_JniStat
+ * Method:    median
+ * Signature: (Ljava/lang/Object;II)D
+ */
+JNIEXPORT jdouble JNICALL Java_org_jetbrains_kotlinx_multik_openblas_stat_JniStat_median
+	(JNIEnv *env, jobject jobj, jobject jarr, jint size, jint type) {
+
+  void *varr = env->GetPrimitiveArrayCritical((jarray)jarr, nullptr);
+
+  double ret = array_median(varr, size, type);
+
+  env->ReleasePrimitiveArrayCritical((jarray)jarr, varr, 0);
+
+  return ret;
+}

multik-openblas/multik_jni/src/main/cpp/jni_JniMath.cpp

@@ -1,8 +1,7 @@
 /*
- * Copyright 2020-2021 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
+ * Copyright 2020-2022 JetBrains s.r.o. Use of this source code is governed by the Apache 2.0 license.
  */
 
-#include <map>
 #include "jni_JniMath.h"
 #include "mk_math.h"