implement sd method

Author: jbytecode

src/main/scala/sd.scala

@@ -0,0 +1,41 @@
+package org.expr.mcdm
+
+case class SdResult(
+    cmins: Vec,
+    cmaxs: Vec,
+    normalizedMat: Mat,
+    sds: Vec,
+    sumsds: Double,
+    weights: Vec
+) extends MCDMResult
+
+def sd(
+    mat: Mat,
+    directions: Array[Direction],
+    normalization: NormalizationFunction =
+      Normalization.MaxMinRangeNormalization,
+    options: Map[String, Any] = Map.empty
+): SdResult =
+
+  val (n, p) = Matrix.size(mat)
+
+  val cmins = Matrix.colmins(mat)
+
+  val cmaxs = Matrix.colmaxs(mat)
+
+  val normalizedMat = normalization(mat, Array.emptyDoubleArray, directions)
+
+  val sds = normalizedMat.transpose.map(Statistics.std(_))
+
+  val sumsds = sds.sum
+
+  val weights = sds.map(_ / sumsds)
+
+  SdResult(
+    cmins,
+    cmaxs,
+    normalizedMat,
+    sds,
+    sumsds,
+    weights
+  )

src/test/scala/testsd.scala

@@ -0,0 +1,49 @@
+import munit.Assertions as A
+
+import org.expr.mcdm.Matrix
+import org.expr.mcdm.Direction._
+import org.expr.mcdm.critic
+import org.expr.mcdm.sd
+
+class TestSd extends munit.FunSuite {
+
+  test("Sd Example - 1") {
+    val mat = Array(
+      Array(391152.0, 251165, 2063102, 912, 18784, 0.009, 0.049, 0.196),
+      Array(181681.0, 118972, 1310114, 525, 12087, 0.009, 0.042, 0.157),
+      Array(156478.0, 105801, 993245, 708, 12279, 0.01, 0.041, 0.177),
+      Array(57145.0, 34707, 339417, 210, 3733, 0.013, 0.055, 0.268),
+      Array(34947.0, 17568, 340159, 77, 2015, 0.014, 0.043, 0.204),
+      Array(32667.0, 19308, 201372, 48, 1091, 0.008, 0.029, 0.217),
+      Array(28945.0, 18033, 117762, 48, 886, 0.007, 0.021, 0.178),
+      Array(18893.0, 13816, 139431, 35, 943, 0.01, 0.035, 0.213),
+      Array(18191.0, 9088, 47664, 43, 731, 0.01, 0.021, 0.186),
+      Array(12852.0, 4185, 64770, 3, 376, 0.011, 0.075, 0.285),
+      Array(10878.0, 7107, 11200, 1, 78, 0.003, 0.033, 0.198),
+      Array(4958.0, 1730, 4656, 7, 274, 0.017, 0.053, 0.215),
+      Array(3901.0, 2318, 15598, 17, 357, 0.023, 0.001, 0.155),
+      Array(2742.0, 1042, 52632, 1, 106, 0.022, 0.1, 0.384),
+      Array(1734.0, 771, 1894, 1, 33, 0.011, 0.125, 0.709),
+      Array(1677.0, 568, 1941, 1, 39, 0.011, 0.129, 0.633)
+    )
+
+    val directions = Array(
+      Maximize,
+      Maximize,
+      Maximize,
+      Minimize,
+      Minimize,
+      Minimize,
+      Maximize,
+      Maximize
+    )
+
+    val result = sd(mat, directions)
+
+    val expectedWeights =
+      Array(0.1161, 0.1175, 0.1254, 0.1376, 0.1334, 0.1136, 0.1251, 0.1308)
+
+    A.assert(Matrix.elementwise_equal(result.weights, expectedWeights, 1e-03))
+  }
+
+}