46 | * The algorithm has the following pros over other approaches. The results are
47 | * not highly affected by the parameters and it provides an automatic
48 | * statistically intuitive cut off to determine the outliers.
49 | *
50 | *
51 | * The computation of the LOCI requires the calculation of MDEF and σMDEF.
52 | * MDEF for a point pi at radius r refers to the deviation of the density of pi
53 | * to that in its average local neighborhood density. σMDEF is the
54 | * normalised standard deviation of the point relative to its local
55 | * neighborhood.
56 | *
57 | *
58 | * The original publication suggests the following flagging scheme the object
59 | * should be flagged as an outlier if MDEF(pi, r, α) > 3
60 | * *σMDEF(pi, r, &alpha). The operator produces an outlier score which
61 | * corresponds to the maximum ratio between MDEF(pi, r, α) and
62 | * &sigmaMDEF(pi, r, &alpha) over all r. The higher the ratio the more outlier
63 | * the object is. The proposed threshold to determine outliers is 3.
64 | *
65 | *
66 | * @author Mennatallah Amer
67 | *
68 | */
69 | public class LOCIAnomalyDetectionOperator extends
70 | AbstractNearestNeighborBasedAnomalyDetectionOperator {
71 |
72 | /**
73 | * The parameter name for " The minimum number of neighbors in the
74 | * sampling neighborhood. "
75 | **/
76 | public static String PARAMETER_N_MIN = "n min";
77 |
78 | /**
79 | * The parameter name for " The ratio of the counting neighborhood
80 | * radius to the sampling neighborhood radius. "
81 | **/
82 | public static String PARAMETER_ALPHA = "alpha";
83 |
84 | private DistanceMeasureHelper measureHelper = new DistanceMeasureHelper(
85 | this);
86 |
87 | public LOCIAnomalyDetectionOperator(OperatorDescription description) {
88 | super(description);
89 | getExampleSetInput().addPrecondition(
90 | new DistanceMeasurePrecondition(getExampleSetInput(), this));
91 | }
92 |
93 | @Override
94 | public double[] doWork(ExampleSet exampleSet, Attributes attributes,
95 | double[][] points, int[] weight) throws OperatorException {
96 | DistanceMeasure measure = measureHelper
97 | .getInitializedMeasure(exampleSet);
98 | double alpha = getParameterAsDouble(PARAMETER_ALPHA);
99 | int nmin = getParameterAsInt(PARAMETER_N_MIN);
100 | int n = points.length;
101 | double[] ret = {1};
102 |
103 | if (n > 1) {
104 | if (nmin == n) {
105 | this.logWarning("Setting " + PARAMETER_N_MIN + " to #Datapoints-1 because n min can't be equal #Datapoints.");
106 | nmin = n-1;
107 | //this.setParameter(PARAMETER_N_MIN, (n-1)+"");
108 | }
109 |
110 | LOCIEvaluator evaluator = new LOCIEvaluator(measure, alpha, nmin,
111 | points, weight);
112 | ret = evaluator.evaluate();
113 | }
114 | return ret;
115 | }
116 |
117 | public List getParameterTypes() {
118 | LinkedList types = new LinkedList();
119 | types
120 | .add(new ParameterTypeDouble(
121 | PARAMETER_ALPHA,
122 | "The ratio of the counting neighborhood radius to the sampling neighborhood radius.",
123 | 0, 1, 0.5));
124 | types
125 | .add(new ParameterTypeInt(
126 | PARAMETER_N_MIN,
127 | "The minimum number of neighbors in the sampling neighborhood.",
128 | 1, Integer.MAX_VALUE, 20, false));
129 |
130 | types.addAll(DistanceMeasures.getParameterTypes(this));
131 | return types;
132 |
133 | }
134 |
135 | }
136 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/nearest_neighbor_based/LOFAnomalyDetectionOperator.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2011 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see
40 | * The LOF anomaly detection calculates the anomaly score according to the local
41 | * outlier factor algorithm proposed by Breunig et al[1999;2000].
42 | *
43 | *
44 | *
45 | * LOF is one of the earliest local density based approaches proposed. There are
46 | * several steps in the calculation of the LOF. The initial step involves
47 | * getting the nearest neighbors set.The definition of the k-distance employed
48 | * is the one proposed in the original paper in order to handle duplicates. The
49 | * definition states that the k-distance(p) has at least k neighbors with
50 | * distinct spatial coordinates that have a distance less than or equal it and
51 | * at most k-1 of such neighbors with distance strictly less than it. The
52 | * reachability distance (reach-dist(p,o)) is the maximum of the distance
53 | * between point p and o and the k-distance(o). The local reachability is the
54 | * inverse of the average reachability distance over the nearest neighborhood
55 | * set. Finally the LOF is calculated as the average of the ratio of the local
56 | * reachability density over the neighborhood set. The values of the LOF
57 | * oscillates with the change in the size of the neighborhood. Thus a range is
58 | * defined for the size of the neighborhood. The maximum LOF over that range is
59 | * taken as the final LOF score.
60 | *
61 | *
62 | * A normal instance has an outlier value of approximately 1, while outliers
63 | * have values greater than 1.
64 | *
65 | *
66 | * @author Mennatallah Amer
67 | *
68 | */
69 | public class LOFAnomalyDetectionOperator extends KNNAnomalyDetectionOperator {
70 |
71 | public static String PARAMETER_MINIMUM_K = "k_min (MinPtsLB)";
72 | public static String PARAMETER_MAXIMUM_K = "k_max (MinPtsUB)";
73 |
74 | public LOFAnomalyDetectionOperator(OperatorDescription description) {
75 | super(description);
76 | }
77 |
78 | @Override
79 | public double[] doWork(ExampleSet exampleSet, Attributes attributes,
80 | double[][] points, int[] weight) throws OperatorException {
81 | DistanceMeasure measure = getMeasureHelper().getInitializedMeasure(
82 | exampleSet);
83 |
84 | int n = points.length;
85 | int minK = getParameterAsInt(PARAMETER_MINIMUM_K);
86 | int maxK = getParameterAsInt(PARAMETER_MAXIMUM_K);
87 | double[] ret = {1};
88 |
89 | if (n > 1) {
90 | if (maxK >= n) {
91 | this.logWarning("Setting " + PARAMETER_MAXIMUM_K + " to "+ (n-1) + " because there cannot be more neighbors than data points.");
92 | maxK = n-1;
93 | //this.setParameter(PARAMETER_MAXIMUM_K, maxK+"");
94 | }
95 | if (maxK < minK) {
96 | this.logWarning("Setting " + PARAMETER_MINIMUM_K + " to "+ maxK + " to make UpperBound at least as large as LowerBound.");
97 | minK = maxK;
98 | //this.setParameter(PARAMETER_MINIMUM_K, minK+"");
99 | }
100 | boolean parallel = getParameterAsBoolean(PARAMETER_PARALLELIZE_EVALUATION_PROCESS);
101 | int numberOfThreads = getParameterAsInt(PARAMETER_NUMBER_OF_THREADS);
102 |
103 | readModel(n,maxK,points,weight,measure);
104 | //KNNCollection knnCollection = new KNNCollection(n, maxK, points, weight);
105 | LOFEvaluator evaluator = new LOFEvaluator(minK, knnCollection,
106 | measure,parallel, numberOfThreads, this, n, maxK , newCollection);
107 | ret = evaluator.evaluate();
108 | if(newCollection) {
109 | model = new KNNCollectionModel(exampleSet,knnCollection,measure);
110 | }
111 | else {
112 | model = new KNNCollectionModel(exampleSet,modelInput.getData(KNNCollectionModel.class).get(),measure);
113 | }
114 | modelOutput.deliver(model);
115 | knnCollection = null;
116 |
117 | }
118 | return ret;
119 | }
120 |
121 | @Override
122 | public List getParameterTypes() {
123 | List types = super.getParameterTypes();
124 | types.get(0).setKey(PARAMETER_MINIMUM_K);
125 | types.get(0).setDescription("The lower bound of MinPts");
126 | types.remove(1);
127 |
128 | types.add(1, new ParameterTypeInt(PARAMETER_MAXIMUM_K,
129 | "The upper bound of the MinPts ", 1, Integer.MAX_VALUE, 20,
130 | false));
131 |
132 | return types;
133 | }
134 |
135 | }
136 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/nearest_neighbor_based/LoOPAnomalyDetectionOperator.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2011 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see 1) {
85 | if (k >= n) {
86 | this.logWarning("Setting " + KNNAnomalyDetectionOperator.PARAMETER_K + " to #Datapoints-1.");
87 | k = n-1;
88 | //this.setParameter(KNNAnomalyDetectionOperator.PARAMETER_K, (n-1)+"");
89 | }
90 | boolean parallel = getParameterAsBoolean(PARAMETER_PARALLELIZE_EVALUATION_PROCESS);
91 | int numberOfThreads = getParameterAsInt(PARAMETER_NUMBER_OF_THREADS);
92 |
93 | readModel(n,k,points,weight,measure);
94 | LoOPEvaluator evaluator = new LoOPEvaluator(knnCollection,
95 | measure, lamda,parallel, numberOfThreads, this,n,k,newCollection);
96 |
97 | ret = evaluator.evaluate();
98 | model = new KNNCollectionModel(exampleSet,knnCollection,measure);
99 | modelOutput.deliver(model);
100 | knnCollection = null;
101 | }
102 | return ret;
103 | }
104 |
105 | public List getParameterTypes() {
106 | List types = super.getParameterTypes();
107 | types.remove(1);
108 | types
109 | .add(
110 | 1,
111 | new ParameterTypeDouble(
112 | PARAMETER_LAMBDA,
113 | "The normalization factor. The results are weakly affected by this factor. ",
114 | 1, 3, 3, true));
115 |
116 | return types;
117 | }
118 |
119 | }
120 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/nearest_neighbor_based/Point.java:
--------------------------------------------------------------------------------
1 | package de.dfki.madm.anomalydetection.operator.nearest_neighbor_based;
2 |
3 |
4 |
5 | public class Point implements Comparable {
6 | public int getIndex() {
7 | return index;
8 | }
9 |
10 | double[] point;
11 | int index;
12 |
13 | public Point(int index, double[] point) {
14 | this.index = index;
15 | this.point = point;
16 | }
17 |
18 | @Override
19 | public int compareTo(Point arg0) {
20 | int n = point.length;
21 | for (int i = 0; i < n; i++)
22 | if (point[i] != arg0.point[i])
23 | if (point[i] < arg0.point[i])
24 | return -1;
25 | else
26 | return 1;
27 |
28 | return 0;
29 | }
30 |
31 | public boolean equals(Object obj) {
32 |
33 | return compareTo((Point) obj) == 0;
34 | }
35 |
36 | }
37 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/statistical_based/OutlierCellColorProvider.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2014 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see colors;
34 | private int numberOfSpecialAttributes;
35 | private int color;
36 | private OutlierExampleSet exampleSet;
37 | private DataViewerTable dataTable;
38 |
39 | public OutlierCellColorProvider(DataViewerTable dataTable,HashMap colors,int numberOfSpecialAttributes,OutlierExampleSet exampleSet) {
40 | this.colors = colors;
41 | this.numberOfSpecialAttributes = numberOfSpecialAttributes;
42 | this.dataTable = dataTable;
43 | this.exampleSet = exampleSet;
44 | }
45 |
46 | @Override
47 | public Color getCellColor(int row, int column) {
48 | int col = dataTable.convertColumnIndexToModel(column);
49 | if(col <= numberOfSpecialAttributes) {
50 | if(col == 0) {
51 | // first column (index)
52 | if (row % 2 == 0) {
53 | return Color.WHITE;
54 | } else {
55 | return SwingTools.LIGHTEST_BLUE;
56 | }
57 | }
58 | if (row % 2 == 0) {
59 | //i special attributes (like score)
60 | return Color.WHITE;
61 | } else {
62 | return SwingTools.LIGHTEST_YELLOW;
63 | }
64 | }
65 | else {
66 | //i row index
67 | String index = dataTable.getCell(row+1,0);
68 | // attribute name
69 | String attribute_name = dataTable.getCell(0,column);
70 | // index of that attribute needed to fetch the right color value
71 | int tableIndex = exampleSet.getAttributes().get(attribute_name).getTableIndex();
72 | color = (int)colors.get(tableIndex)[Integer.parseInt(index)-1];
73 |
74 | /*transform the integer values into actual colors.
75 | * 510 => green (0,255,0)
76 | * 255 => yellow (255,255,0)
77 | * 0 => red (255,0,0)
78 | * ...
79 | */
80 | if(color == -1){
81 | return Color.white;
82 | }
83 | else if(color >= 255) {
84 | return new Color((510-color), 255, 0);
85 | }
86 | // second half of colors go from yellow to red (unnormal bins)
87 | else{
88 | return new Color(255,(color) , 0);
89 | }
90 | }
91 |
92 | }
93 |
94 | }
95 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/statistical_based/OutlierColorJoin.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2014 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see eColor = new HashMap();
69 | for(Attribute att : exampleSetAttributes) {
70 | int[] colors = new int[exampleSet.size()];
71 | Arrays.fill(colors,-1);
72 | eColor.put(att.getTableIndex(), colors);
73 | }
74 | //for(int x = 0; x< exampleSet.size();x++) {
75 | int x = 0;
76 | outlierExampleSet.remapIds();
77 |
78 | for(Example e : exampleSet) {
79 | currentId = e.getId();
80 | int[] i = outlierExampleSet.getExampleIndicesFromId(currentId);
81 | try {
82 | if(i.length > 1 ) {
83 | throw new OperatorException("Ids are not unique.");
84 | }
85 | }
86 | catch (NullPointerException E){
87 | this.logNote("Data Row with id " +currentId+ " not found in the OutlierExampleSet");
88 | }
89 | for(Attribute att: exampleSetAttributes) {
90 | for(Attribute outAtt : outlierExampleSetAttributes){
91 | // find attribute with the same name
92 | if(att.getName() == outAtt.getName()) {
93 | //save color
94 | int[] colors = eColor.get(att.getTableIndex());
95 |
96 | try {
97 | colors[x] = outlierExampleSet.colors.get(att.getTableIndex())[i[0]];
98 | }
99 | catch (NullPointerException E ){
100 | colors[x] = -1;
101 | this.logNote("Attribute "+ att.getName() + " not found in the OutlierExampleSet");
102 | }
103 | eColor.put(att.getTableIndex(), colors);
104 | }
105 | }
106 |
107 | }
108 | x++;
109 | }
110 | return new OutlierExampleSet(exampleSet,eColor);
111 |
112 | }
113 | }
114 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/statistical_based/OutlierDataViewer.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2014 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see applicableFilterNames = new LinkedList();
94 | for (String conditionName : ConditionedExampleSet.KNOWN_CONDITION_NAMES) {
95 | try {
96 | ConditionedExampleSet.createCondition(conditionName, exampleSet, null);
97 | applicableFilterNames.add(conditionName);
98 | } catch (ConditionCreationException ex) {} // Do nothing
99 | }
100 | String[] applicableConditions = new String[applicableFilterNames.size()];
101 | applicableFilterNames.toArray(applicableConditions);
102 | final JComboBox filterSelector = new JComboBox(applicableConditions);
103 | filterSelector.setToolTipText("These filters can be used to skip examples in the view fulfilling the filter condition.");
104 | filterSelector.addItemListener(new ItemListener() {
105 | @Override
106 | public void itemStateChanged(ItemEvent e) {
107 | updateFilter((String)filterSelector.getSelectedItem());
108 | }
109 | });
110 |
111 | int maxNumberBeforeFiltering = DEFAULT_MAX_SIZE_FOR_FILTERING;
112 | String maxString = ParameterService.getParameterValue(RapidMinerGUI.PROPERTY_RAPIDMINER_GUI_MAX_STATISTICS_ROWS);
113 | if (maxString != null) {
114 | try {
115 | maxNumberBeforeFiltering = Integer.parseInt(maxString);
116 | } catch (NumberFormatException e) {
117 | // do nothing
118 | }
119 | }
120 | if (exampleSet.size() > maxNumberBeforeFiltering) {
121 | filterSelector.setEnabled(false);
122 | }
123 | toolBar.add(filterSelector, ViewToolBar.RIGHT);
124 | toolBar.setPreferredSize(new Dimension(getWidth(), 29));
125 | }
126 |
127 | add(toolBar, BorderLayout.NORTH);
128 | JScrollPane tableScrollPane = new ExtendedJScrollPane(dataTable);
129 | tableScrollPane.setBorder(null);
130 | add(tableScrollPane, BorderLayout.CENTER);
131 | setExampleSet(exampleSet);
132 |
133 | /*
134 | * set the right cell color Proiver
135 | */
136 | final int numberOfSpecialAttributes = exampleSet.getAttributes().specialSize();
137 | dataTable.setCellColorProvider(new OutlierCellColorProvider(dataTable,exampleSet.colors,numberOfSpecialAttributes, exampleSet));
138 | }
139 |
140 | public void setExampleSet(ExampleSet exampleSet) {
141 | dataTable.setExampleSet(exampleSet);
142 | }
143 |
144 | private void updateFilter(String conditionName) {
145 | ExampleSet filteredExampleSet = originalExampleSet;
146 | try {
147 | Condition condition = ConditionedExampleSet.createCondition(conditionName, originalExampleSet, null);
148 | filteredExampleSet = new ConditionedExampleSet(originalExampleSet, condition);
149 | } catch (ConditionCreationException ex) {
150 | originalExampleSet.getLog().logError("Cannot create condition '" + conditionName + "' for filtered data view: " + ex.getMessage() + ". Using original data set view...");
151 | filteredExampleSet = originalExampleSet;
152 | }
153 | updateFilterCounter(filteredExampleSet);
154 | setExampleSet(filteredExampleSet);
155 | }
156 |
157 | private void updateFilterCounter(ExampleSet filteredExampleSet) {
158 | filterCounter.setText("(" + filteredExampleSet.size() + " / " + originalExampleSet.size() + "): ");
159 | }
160 |
161 | @Override
162 | public void prepareReporting() {
163 | dataTable.prepareReporting();
164 | }
165 |
166 | @Override
167 | public void finishReporting() {
168 | dataTable.finishReporting();
169 | }
170 |
171 | @Override
172 | public String getColumnName(int columnIndex) {
173 | return dataTable.getColumnName(columnIndex);
174 | }
175 |
176 | @Override
177 | public String getCell(int row, int column) {
178 | return dataTable.getCell(row, column);
179 | }
180 |
181 | @Override
182 | public int getColumnNumber() {
183 | return dataTable.getColumnNumber();
184 | }
185 |
186 | @Override
187 | public int getRowNumber() {
188 | return dataTable.getRowNumber();
189 | }
190 |
191 | @Override
192 | public boolean isFirstLineHeader() { return false; }
193 |
194 | @Override
195 | public boolean isFirstColumnHeader() { return false; }
196 |
197 |
198 |
199 |
200 | }
201 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/statistical_based/OutlierExampleSet.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2014 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see colors;
38 | public OutlierExampleSet(ExampleSet exampleSet,HashMap colors){
39 | this.colors = colors;
40 | this.data = exampleSet;
41 | }
42 | @Override
43 | public String getName() {
44 | return "OutlierExampleSet";
45 | }
46 | @Override
47 | public Attributes getAttributes() {
48 | return data.getAttributes();
49 | }
50 |
51 | @Override
52 | public Example getExample(int arg0) {
53 | return data.getExample(arg0);
54 | }
55 |
56 | @Override
57 | public ExampleTable getExampleTable() {
58 | return data.getExampleTable();
59 | }
60 |
61 | @Override
62 | public int size() {
63 | return data.size();
64 | }
65 | @Override
66 | public OutlierExampleSet clone() {
67 | OutlierExampleSet result = new OutlierExampleSet((ExampleSet)data.clone(),(HashMap)colors.clone());
68 | return result;
69 |
70 | }
71 | @Override
72 | public Iterator iterator() {
73 | return data.iterator();
74 | }
75 |
76 |
77 | }
78 |
--------------------------------------------------------------------------------
/src/de/dfki/madm/anomalydetection/operator/statistical_based/OutlierExampleSetDataRenderer.java:
--------------------------------------------------------------------------------
1 | /*
2 | * RapidMiner Anomaly Detection Extension
3 | *
4 | * Copyright (C) 2009-2014 by Deutsches Forschungszentrum fuer
5 | * Kuenstliche Intelligenz GmbH or its licensors, as applicable.
6 | *
7 | * This is free software: you can redistribute it and/or modify
8 | * it under the terms of the GNU Affero General Public License as published by
9 | * the Free Software Foundation, either version 3 of the License, or
10 | * (at your option) any later version.
11 | *
12 | * You should have received a copy of the GNU Affero General Public License
13 | * along with this software. If not, see 
47 | [](https://www.goldiges.de/)
48 |
--------------------------------------------------------------------------------
/build.properties:
--------------------------------------------------------------------------------
1 | extension.version=2
2 | extension.revision=4
3 | extension.update=001
4 |
--------------------------------------------------------------------------------
/build.xml:
--------------------------------------------------------------------------------
1 |