implement piv

Author: jbytecode

src/main/scala/piv.scala

@@ -0,0 +1,44 @@
+package org.expr.mcdm
+
+import org.expr.mcdm.Direction.{Minimize, Maximize}
+
+case class PivResult(
+    scores: Vec
+) extends MCDMResult
+
+def piv(
+    decmat: Mat,
+    weights: Vec,
+    directions: Array[Direction],
+    normalization: NormalizationFunction =
+      Normalization.VectorNormNormalization,
+    options: Map[String, Any] = Map.empty 
+): PivResult = 
+
+    val (nrow, ncol) = Matrix.size(decmat)
+
+    val normalized_dec_mat = normalization(decmat, weights, directions)
+
+    val weighted_norm_mat = Matrix.weightizeColumns(normalized_dec_mat, weights)
+
+    val dirfunctions = directions.map {
+      case Maximize => Statistics.maximum
+      case Minimize => Statistics.minimum
+    }
+    val desiredvalues = Matrix.applyFunctionsToColumns(weighted_norm_mat, dirfunctions)
+
+    var finalmat = Matrix.zeros(nrow, ncol)
+
+    for i <- 0 until nrow do
+      for j <- 0 until ncol do
+        if directions(j) == Maximize then
+          finalmat(i)(j) = desiredvalues(j) - weighted_norm_mat(i)(j)
+        else if directions(j) == Minimize then
+          finalmat(i)(j) = weighted_norm_mat(i)(j) - desiredvalues(j)
+
+    // di values are scores
+    val di = Matrix.rowsums(finalmat)
+
+    PivResult(
+        di
+    )

src/test/scala/testpiv.scala

@@ -0,0 +1,32 @@
+import munit.Assertions as A
+
+import org.expr.mcdm.piv
+import org.expr.mcdm.Direction.{Maximize, Minimize}
+import org.expr.mcdm.Matrix
+
+class TestPiv extends munit.FunSuite {
+  test("Piv Example - 1") {
+
+    val decmat = Array(
+      Array(60.0, 2.5, 2540.0, 500.0, 990.0),
+      Array(6.35, 6.667, 1016.0, 3000.0, 1041.0),
+      Array(6.8, 10.0, 1727.0, 1500.0, 1676.0),
+      Array(10.0, 5.0, 1000.0, 2000.0, 965.0),
+      Array(2.5, 9.8, 560.0, 500.0, 915.0),
+      Array(4.5, 12.5, 1016.0, 350.0, 508.0),
+      Array(3.0, 10.0, 1778.0, 1000.0, 920.0)
+    )
+
+    val weights = Array(0.1761, 0.2042, 0.2668, 0.1243, 0.2286)
+
+    val directions = Array(Maximize, Maximize, Maximize, Maximize, Maximize)
+
+    val expected_scores = Array(0.22086675609968962, 0.35854940101222144,
+      0.2734184099704686, 0.4005382183676046, 0.4581157878699193,
+      0.43595371718873477, 0.3580651803072459)
+
+    val result = piv(decmat, weights, directions)
+
+    A.assert(Matrix.elementwise_equal(result.scores, expected_scores, 1e-05), "PIV scores do not match expected values")
+  }
+}