feat: add analyzer to order qi

examples/order.json

@@ -0,0 +1,15 @@
+{"x":4, "y":1}
+{"x":4, "y":2}
+{"x":4, "y":3}
+{"x":3, "y":4}
+{"x":3, "y":5}
+{"x":3, "y":6}
+{"x":3, "y":7}
+{"x":4, "y":8}
+{"x":4, "y":9}
+{"x":4, "y":10}
+{"x":3, "y":11}
+{"x":4, "y":12}
+{"x":4, "y":13}
+{"x":3, "y":14}
+{"x":3, "y":15}

pkg/sigo/analyzer.go

@@ -0,0 +1,70 @@
+package sigo
+
+import (
+	"sort"
+)
+
+type Source struct {
+	qi     map[string]int
+	values map[string][]float64
+}
+
+func NewAnalyzer(qi []string) Analyzer {
+	dict := make(map[string]int)
+	for i, key := range qi {
+		dict[key] = i
+	}
+
+	return Source{qi: dict, values: make(map[string][]float64)}
+}
+
+func (s Source) Add(r Record) {
+	for key, i := range s.qi {
+		s.values[key] = append(s.values[key], r.QuasiIdentifer()[i])
+	}
+}
+
+func (s Source) QI(i int) string {
+	return s.Order()[i]
+}
+
+func (s Source) CountUniqueValues() map[string]int {
+	uniques := make(map[string]int)
+
+	for key := range s.qi {
+		uniques[key] = Unique(s.values[key])
+	}
+
+	return uniques
+}
+
+func (s Source) Order() map[int]string {
+	order := make(map[int]string)
+	switched := make(map[int][]string)
+	slice := []int{}
+
+	for key, count := range s.CountUniqueValues() {
+		switched[count] = append(switched[count], key)
+
+		slice = append(slice, count)
+	}
+
+	sort.Sort(sort.Reverse(sort.IntSlice(slice)))
+
+	i := 0
+
+	for _, count := range slice {
+		for _, qi := range switched[count] {
+			order[i] = qi
+			i++
+		}
+
+		delete(switched, count)
+	}
+
+	return order
+}
+
+func (s Source) Dimension(rot int) int {
+	return s.qi[s.Order()[rot]]
+}

pkg/sigo/analyzer_test.go

@@ -0,0 +1,71 @@
+// Copyright (C) 2022 CGI France
+//
+// This file is part of SIGO.
+//
+// SIGO is free software: you can redistribute it and/or modify
+// it under the terms of the GNU General Public License as published by
+// the Free Software Foundation, either version 3 of the License, or
+// (at your option) any later version.
+//
+// SIGO is distributed in the hope that it will be useful,
+// but WITHOUT ANY WARRANTY; without even the implied warranty of
+// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+// GNU General Public License for more details.
+//
+// You should have received a copy of the GNU General Public License
+// along with SIGO.  If not, see <http://www.gnu.org/licenses/>.
+package sigo_test
+
+import (
+	"testing"
+
+	"github.com/cgi-fr/sigo/pkg/sigo"
+	"github.com/stretchr/testify/assert"
+)
+
+func TestCountUniqueValues(t *testing.T) {
+	t.Parallel()
+
+	qi := []string{"x", "y"}
+	source := sigo.NewAnalyzer(qi)
+
+	source.Add(createRow(4, 1, qi))
+	source.Add(createRow(3, 2, qi))
+	source.Add(createRow(4, 3, qi))
+
+	res := source.CountUniqueValues()
+
+	assert.Equal(t, 2, res["x"])
+	assert.Equal(t, 3, res["y"])
+}
+
+func TestOrderMap(t *testing.T) {
+	t.Parallel()
+
+	qi := []string{"x", "y"}
+	source := sigo.NewAnalyzer(qi)
+
+	source.Add(createRow(4, 1, qi))
+	source.Add(createRow(3, 2, qi))
+	source.Add(createRow(4, 3, qi))
+
+	res := source.Order()
+
+	assert.Equal(t, "y", res[0])
+	assert.Equal(t, "x", res[1])
+}
+
+func TestDimension(t *testing.T) {
+	t.Parallel()
+
+	qi := []string{"x", "y"}
+	source := sigo.NewAnalyzer(qi)
+
+	source.Add(createRow(4, 1, qi))
+	source.Add(createRow(3, 2, qi))
+	source.Add(createRow(4, 3, qi))
+
+	res := source.Dimension(0)
+
+	assert.Equal(t, 1, res)
+}

pkg/sigo/kdtree.go

@@ -35,7 +35,7 @@ type KDTreeFactory struct{}
 
 func (f KDTreeFactory) New(k int, l int, dim int, qi []string) Generalizer {
 	// nolint: exhaustivestruct
-	tree := KDTree{k: k, l: l, dim: dim, clusterID: make(map[string]int), qi: qi}
+	tree := KDTree{k: k, l: l, dim: dim, clusterID: make(map[string]int), analyzer: NewAnalyzer(qi)}
 	root := NewNode(&tree, "root", 0)
 	root.validate()
 	tree.root = &root
@@ -49,7 +49,7 @@ type KDTree struct {
 	root      *node
 	dim       int
 	clusterID map[string]int
-	qi        []string
+	analyzer  Analyzer
 }
 
 func NewKDTree(k, l, dim int, clusterID map[string]int) KDTree {
@@ -59,6 +59,7 @@ func NewKDTree(k, l, dim int, clusterID map[string]int) KDTree {
 
 func (t KDTree) Add(r Record) {
 	t.root.Add(r)
+	t.analyzer.Add(r)
 }
 
 func (t KDTree) Build() {
@@ -112,7 +113,7 @@ func (n *node) incRot() {
 
 func (n *node) build() {
 	log.Debug().
-		Str("Dimension", n.tree.qi[n.rot]).
+		Str("Dimension", n.tree.analyzer.QI(n.rot)).
 		Str("Path", n.clusterPath).
 		Int("Size", len(n.cluster)).
 		Msg("Cluster:")
@@ -174,8 +175,9 @@ func (n *node) Bounds() []bounds {
 }
 
 func (n *node) split() (node, node, bool) {
+	dim := n.tree.analyzer.Dimension(n.rot)
 	sort.SliceStable(n.cluster, func(i int, j int) bool {
-		return n.cluster[i].QuasiIdentifer()[n.rot] < n.cluster[j].QuasiIdentifer()[n.rot]
+		return n.cluster[i].QuasiIdentifer()[dim] < n.cluster[j].QuasiIdentifer()[dim]
 	})
 
 	n.pivot = nil
@@ -189,7 +191,7 @@ func (n *node) split() (node, node, bool) {
 	previous := n.cluster[0]
 
 	for _, row := range n.cluster {
-		if lowerSize < len(n.cluster)/2 || row.QuasiIdentifer()[n.rot] == previous.QuasiIdentifer()[n.rot] {
+		if lowerSize < len(n.cluster)/2 { // || row.QuasiIdentifer()[n.rot] == previous.QuasiIdentifer()[n.rot] {
 			lower.Add(row)
 			previous = row
 			lowerSize++

pkg/sigo/model.go

@@ -59,3 +59,11 @@ type Anonymizer interface {
 type Debugger interface {
 	Information(Record, Cluster) Record
 }
+
+type Analyzer interface {
+	Add(Record)
+	QI(i int) string
+	CountUniqueValues() map[string]int
+	Order() map[int]string
+	Dimension(int) int
+}

pkg/sigo/stats.go

@@ -210,3 +210,13 @@ func BoxMuller() (float64, float64) {
 
 	return z1, z2
 }
+
+func Unique(values []float64) int {
+	tmp := make(map[float64]int)
+
+	for _, val := range values {
+		tmp[val]++
+	}
+
+	return len(tmp)
+}

pkg/sigo/stats_test.go

@@ -32,3 +32,16 @@ func TestQuartiles(t *testing.T) {
 	assert.Equal(t, q.Q2, sigo.Median(values))
 	assert.Equal(t, 5.00, sigo.IQR(values))
 }
+
+func TestUnique(t *testing.T) {
+	t.Parallel()
+
+	values1 := []float64{12, 10, 5, 6, 9, 10, 4, 5, 10, 12, 9, 6, 4, 3, 9, 10}
+	values2 := []float64{1, 9, 8, 5, 2, 6, 7, 10, 3, 12, 4, 11}
+
+	res1 := sigo.Unique(values1)
+	res2 := sigo.Unique(values2)
+
+	assert.Equal(t, 7, res1)
+	assert.Equal(t, 12, res2)
+}

test/suites/02-order-QI.yml

@@ -0,0 +1,30 @@
+# Venom Test Suite definition
+# Check Venom documentation for more information : https://github.com/ovh/venom
+
+name: sigo odrer qi
+testcases:
+  - name: sort qi
+    steps:
+      - script: |-
+          sigo -q x,y -i id <<EOF
+          {"x":4, "y":1}
+          {"x":4, "y":2}
+          {"x":4, "y":3}
+          {"x":3, "y":4}
+          {"x":3, "y":5}
+          {"x":3, "y":6}
+          {"x":3, "y":7}
+          {"x":4, "y":8}
+          {"x":4, "y":9}
+          {"x":4, "y":10}
+          {"x":3, "y":11}
+          {"x":4, "y":12}
+          {"x":4, "y":13}
+          {"x":3, "y":14}
+          {"x":3, "y":15}
+          EOF
+        assertions:
+          - result.systemoutjson.id ShouldEqual 1
+          - result.systemoutjson.x ShouldEqual 3
+          - result.systemoutjson.y ShouldEqual 4
+          - result.code ShouldEqual 0