nlptoolkit-classification 1.0.8 → 1.0.9

package/dist/Model/NeuralNetwork/MultiLayerPerceptronModel.js

@@ -1,127 +1,119 @@
-(function (factory) {
-    if (typeof module === "object" && typeof module.exports === "object") {
-        var v = factory(require, exports);
-        if (v !== undefined) module.exports = v;
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.MultiLayerPerceptronModel = void 0;
+const LinearPerceptronModel_1 = require("./LinearPerceptronModel");
+const Matrix_1 = require("nlptoolkit-math/dist/Matrix");
+const ActivationFunction_1 = require("../../Parameter/ActivationFunction");
+const ClassificationPerformance_1 = require("../../Performance/ClassificationPerformance");
+const Vector_1 = require("nlptoolkit-math/dist/Vector");
+const Random_1 = require("nlptoolkit-util/dist/Random");
+const FileContents_1 = require("nlptoolkit-util/dist/FileContents");
+const Partition_1 = require("../../InstanceList/Partition");
+class MultiLayerPerceptronModel extends LinearPerceptronModel_1.LinearPerceptronModel {
+    V;
+    activationFunction;
+    /**
+     * The allocateWeights method allocates layers' weights of Matrix W and V.
+     *
+     * @param H Integer value for weights.
+     * @param random Random function to set weights.
+     */
+    allocateWeights(H, random) {
+        this.W = this.allocateLayerWeights(H, this.d + 1, random);
+        this.V = this.allocateLayerWeights(this.K, H + 1, random);
     }
-    else if (typeof define === "function" && define.amd) {
-        define(["require", "exports", "./LinearPerceptronModel", "nlptoolkit-math/dist/Matrix", "../../Parameter/ActivationFunction", "../../Performance/ClassificationPerformance", "nlptoolkit-math/dist/Vector", "nlptoolkit-util/dist/Random", "nlptoolkit-util/dist/FileContents", "../../InstanceList/Partition"], factory);
+    /**
+     * Loads a multi-layer perceptron model from an input model file.
+     * @param fileName Model file name.
+     */
+    constructor2(fileName) {
+        let input = new FileContents_1.FileContents(fileName);
+        this.loadClassLabels(input);
+        this.W = this.loadMatrix(input);
+        this.V = this.loadMatrix(input);
+        this.activationFunction = this.loadActivationFunction(input);
     }
-})(function (require, exports) {
-    "use strict";
-    Object.defineProperty(exports, "__esModule", { value: true });
-    exports.MultiLayerPerceptronModel = void 0;
-    const LinearPerceptronModel_1 = require("./LinearPerceptronModel");
-    const Matrix_1 = require("nlptoolkit-math/dist/Matrix");
-    const ActivationFunction_1 = require("../../Parameter/ActivationFunction");
-    const ClassificationPerformance_1 = require("../../Performance/ClassificationPerformance");
-    const Vector_1 = require("nlptoolkit-math/dist/Vector");
-    const Random_1 = require("nlptoolkit-util/dist/Random");
-    const FileContents_1 = require("nlptoolkit-util/dist/FileContents");
-    const Partition_1 = require("../../InstanceList/Partition");
-    class MultiLayerPerceptronModel extends LinearPerceptronModel_1.LinearPerceptronModel {
-        /**
-         * The allocateWeights method allocates layers' weights of Matrix W and V.
-         *
-         * @param H Integer value for weights.
-         * @param random Random function to set weights.
-         */
-        allocateWeights(H, random) {
-            this.W = this.allocateLayerWeights(H, this.d + 1, random);
-            this.V = this.allocateLayerWeights(this.K, H + 1, random);
-        }
-        /**
-         * Loads a multi-layer perceptron model from an input model file.
-         * @param fileName Model file name.
-         */
-        constructor2(fileName) {
-            let input = new FileContents_1.FileContents(fileName);
-            this.loadClassLabels(input);
-            this.W = this.loadMatrix(input);
-            this.V = this.loadMatrix(input);
-            this.activationFunction = this.loadActivationFunction(input);
-        }
-        /**
-         * The calculateOutput method calculates the forward single hidden layer by using Matrices W and V.
-         */
-        calculateOutput() {
-            this.calculateForwardSingleHiddenLayer(this.W, this.V, this.activationFunction);
-        }
-        saveTxt(fileName) {
-        }
-        /**
-         * Training algorithm for the multilayer perceptron algorithm. 20 percent of the data is separated as cross-validation
-         * data used for selecting the best weights. 80 percent of the data is used for training the multilayer perceptron with
-         * gradient descent.
-         *
-         * @param train   Training data given to the algorithm
-         * @param params Parameters of the multilayer perceptron.
-         */
-        train(train, params) {
-            this.initialize(train);
-            let parameters = params;
-            let partition = new Partition_1.Partition(train, parameters.getCrossValidationRatio(), true);
-            let trainSet = partition.get(1);
-            let validationSet = partition.get(1);
-            this.activationFunction = parameters.getActivationFunction();
-            this.allocateWeights(parameters.getHiddenNodes(), new Random_1.Random(parameters.getSeed()));
-            let bestW = this.W.clone();
-            let bestV = this.V.clone();
-            let bestClassificationPerformance = new ClassificationPerformance_1.ClassificationPerformance(0.0);
-            let epoch = parameters.getEpoch();
-            let learningRate = parameters.getLearningRate();
-            for (let i = 0; i < epoch; i++) {
-                trainSet.shuffle(new Random_1.Random(parameters.getSeed()));
-                for (let j = 0; j < trainSet.size(); j++) {
-                    this.createInputVector(trainSet.get(j));
-                    let hidden = this.calculateHidden(this.x, this.W, this.activationFunction);
-                    let hiddenBiased = hidden.biased();
-                    let rMinusY = this.calculateRMinusY(trainSet.get(j), hiddenBiased, this.V);
-                    let deltaV = new Matrix_1.Matrix(rMinusY, hiddenBiased);
-                    let tmph = this.V.multiplyWithVectorFromLeft(rMinusY);
-                    tmph.remove(0);
-                    let activationDerivative;
-                    switch (this.activationFunction) {
-                        case ActivationFunction_1.ActivationFunction.SIGMOID:
-                        default:
-                            let oneMinusHidden = this.calculateOneMinusHidden(hidden);
-                            activationDerivative = oneMinusHidden.elementProduct(hidden);
-                            break;
-                        case ActivationFunction_1.ActivationFunction.TANH:
-                            let one = new Vector_1.Vector(hidden.size(), 1.0);
-                            hidden.tanh();
-                            activationDerivative = one.difference(hidden.elementProduct(hidden));
-                            break;
-                        case ActivationFunction_1.ActivationFunction.RELU:
-                            hidden.reluDerivative();
-                            activationDerivative = hidden;
-                            break;
-                    }
-                    let tmpHidden = tmph.elementProduct(activationDerivative);
-                    let deltaW = new Matrix_1.Matrix(tmpHidden, this.x);
-                    deltaV.multiplyWithConstant(learningRate);
-                    this.V.add(deltaV);
-                    deltaW.multiplyWithConstant(learningRate);
-                    this.W.add(deltaW);
-                }
-                let currentClassificationPerformance = this.testClassifier(validationSet);
-                if (currentClassificationPerformance.getAccuracy() > bestClassificationPerformance.getAccuracy()) {
-                    bestClassificationPerformance = currentClassificationPerformance;
-                    bestW = this.W.clone();
-                    bestV = this.V.clone();
+    /**
+     * The calculateOutput method calculates the forward single hidden layer by using Matrices W and V.
+     */
+    calculateOutput() {
+        this.calculateForwardSingleHiddenLayer(this.W, this.V, this.activationFunction);
+    }
+    saveTxt(fileName) {
+    }
+    /**
+     * Training algorithm for the multilayer perceptron algorithm. 20 percent of the data is separated as cross-validation
+     * data used for selecting the best weights. 80 percent of the data is used for training the multilayer perceptron with
+     * gradient descent.
+     *
+     * @param train   Training data given to the algorithm
+     * @param params Parameters of the multilayer perceptron.
+     */
+    train(train, params) {
+        this.initialize(train);
+        let parameters = params;
+        let partition = new Partition_1.Partition(train, parameters.getCrossValidationRatio(), true);
+        let trainSet = partition.get(1);
+        let validationSet = partition.get(1);
+        this.activationFunction = parameters.getActivationFunction();
+        this.allocateWeights(parameters.getHiddenNodes(), new Random_1.Random(parameters.getSeed()));
+        let bestW = this.W.clone();
+        let bestV = this.V.clone();
+        let bestClassificationPerformance = new ClassificationPerformance_1.ClassificationPerformance(0.0);
+        let epoch = parameters.getEpoch();
+        let learningRate = parameters.getLearningRate();
+        for (let i = 0; i < epoch; i++) {
+            trainSet.shuffle(new Random_1.Random(parameters.getSeed()));
+            for (let j = 0; j < trainSet.size(); j++) {
+                this.createInputVector(trainSet.get(j));
+                let hidden = this.calculateHidden(this.x, this.W, this.activationFunction);
+                let hiddenBiased = hidden.biased();
+                let rMinusY = this.calculateRMinusY(trainSet.get(j), hiddenBiased, this.V);
+                let deltaV = new Matrix_1.Matrix(rMinusY, hiddenBiased);
+                let tmph = this.V.multiplyWithVectorFromLeft(rMinusY);
+                tmph.remove(0);
+                let activationDerivative;
+                switch (this.activationFunction) {
+                    case ActivationFunction_1.ActivationFunction.SIGMOID:
+                    default:
+                        let oneMinusHidden = this.calculateOneMinusHidden(hidden);
+                        activationDerivative = oneMinusHidden.elementProduct(hidden);
+                        break;
+                    case ActivationFunction_1.ActivationFunction.TANH:
+                        let one = new Vector_1.Vector(hidden.size(), 1.0);
+                        hidden.tanh();
+                        activationDerivative = one.difference(hidden.elementProduct(hidden));
+                        break;
+                    case ActivationFunction_1.ActivationFunction.RELU:
+                        hidden.reluDerivative();
+                        activationDerivative = hidden;
+                        break;
                 }
-                learningRate *= parameters.getEtaDecrease();
+                let tmpHidden = tmph.elementProduct(activationDerivative);
+                let deltaW = new Matrix_1.Matrix(tmpHidden, this.x);
+                deltaV.multiplyWithConstant(learningRate);
+                this.V.add(deltaV);
+                deltaW.multiplyWithConstant(learningRate);
+                this.W.add(deltaW);
             }
-            this.W = bestW;
-            this.V = bestV;
-        }
-        /**
-         * Loads the multi-layer perceptron model from an input file.
-         * @param fileName File name of the multi-layer perceptron model.
-         */
-        loadModel(fileName) {
-            this.constructor2(fileName);
+            let currentClassificationPerformance = this.testClassifier(validationSet);
+            if (currentClassificationPerformance.getAccuracy() > bestClassificationPerformance.getAccuracy()) {
+                bestClassificationPerformance = currentClassificationPerformance;
+                bestW = this.W.clone();
+                bestV = this.V.clone();
+            }
+            learningRate *= parameters.getEtaDecrease();
         }
+        this.W = bestW;
+        this.V = bestV;
+    }
+    /**
+     * Loads the multi-layer perceptron model from an input file.
+     * @param fileName File name of the multi-layer perceptron model.
+     */
+    loadModel(fileName) {
+        this.constructor2(fileName);
     }
-    exports.MultiLayerPerceptronModel = MultiLayerPerceptronModel;
-});
+}
+exports.MultiLayerPerceptronModel = MultiLayerPerceptronModel;
 //# sourceMappingURL=MultiLayerPerceptronModel.js.map

package/dist/Model/NeuralNetwork/MultiLayerPerceptronModel.js.map

@@ -1 +1 @@
-{"version":3,"file":"MultiLayerPerceptronModel.js","sourceRoot":"","sources":["../../../source/Model/NeuralNetwork/MultiLayerPerceptronModel.ts"],"names":[],"mappings":";;;;;;;;;;;;IAAA,mEAA8D;IAC9D,wDAAmD;IACnD,2EAAsE;IAGtE,2FAAsF;IACtF,wDAAmD;IACnD,wDAAmD;IACnD,oEAA+D;IAE/D,4DAAuD;IAEvD,MAAa,yBAA0B,SAAQ,6CAAqB;QAKhE;;;;;WAKG;QACK,eAAe,CAAC,CAAS,EAAE,MAAc;YAC7C,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,oBAAoB,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC;YAC1D,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,oBAAoB,CAAC,IAAI,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC;QAC9D,CAAC;QAED;;;WAGG;QACH,YAAY,CAAC,QAAgB;YACzB,IAAI,KAAK,GAAG,IAAI,2BAAY,CAAC,QAAQ,CAAC,CAAA;YACtC,IAAI,CAAC,eAAe,CAAC,KAAK,CAAC,CAAA;YAC3B,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;YAC/B,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;YAC/B,IAAI,CAAC,kBAAkB,GAAG,IAAI,CAAC,sBAAsB,CAAC,KAAK,CAAC,CAAA;QAChE,CAAC;QAED;;WAEG;QACO,eAAe;YACrB,IAAI,CAAC,iCAAiC,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,kBAAkB,CAAC,CAAC;QACpF,CAAC;QAED,OAAO,CAAC,QAAgB;QACxB,CAAC;QAED;;;;;;;WAOG;QACH,KAAK,CAAC,KAAmB,EAAE,MAAiB;YACxC,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;YACtB,IAAI,UAAU,GAAoC,MAAO,CAAA;YACzD,IAAI,SAAS,GAAG,IAAI,qBAAS,CAAC,KAAK,EAAE,UAAU,CAAC,uBAAuB,EAAE,EAAE,IAAI,CAAC,CAAC;YACjF,IAAI,QAAQ,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;YAC/B,IAAI,aAAa,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;YACpC,IAAI,CAAC,kBAAkB,GAAG,UAAU,CAAC,qBAAqB,EAAE,CAAC;YAC7D,IAAI,CAAC,eAAe,CAAC,UAAU,CAAC,cAAc,EAAE,EAAE,IAAI,eAAM,CAAC,UAAU,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;YACpF,IAAI,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;YAC3B,IAAI,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;YAC3B,IAAI,6BAA6B,GAAG,IAAI,qDAAyB,CAAC,GAAG,CAAC,CAAC;YACvE,IAAI,KAAK,GAAG,UAAU,CAAC,QAAQ,EAAE,CAAC;YAClC,IAAI,YAAY,GAAG,UAAU,CAAC,eAAe,EAAE,CAAC;YAChD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,EAAE,CAAC,EAAE,EAAE;gBAC5B,QAAQ,CAAC,OAAO,CAAC,IAAI,eAAM,CAAC,UAAU,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;gBACnD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,IAAI,EAAE,EAAE,CAAC,EAAE,EAAE;oBACtC,IAAI,CAAC,iBAAiB,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;oBACxC,IAAI,MAAM,GAAG,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,kBAAkB,CAAC,CAAC;oBAC3E,IAAI,YAAY,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC;oBACnC,IAAI,OAAO,GAAG,IAAI,CAAC,gBAAgB,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC;oBAC3E,IAAI,MAAM,GAAG,IAAI,eAAM,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;oBAC/C,IAAI,IAAI,GAAG,IAAI,CAAC,CAAC,CAAC,0BAA0B,CAAC,OAAO,CAAC,CAAC;oBACtD,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;oBACf,IAAI,oBAAoB,CAAA;oBACxB,QAAQ,IAAI,CAAC,kBAAkB,EAAC;wBAC5B,KAAK,uCAAkB,CAAC,OAAO,CAAC;wBAChC;4BACI,IAAI,cAAc,GAAG,IAAI,CAAC,uBAAuB,CAAC,MAAM,CAAC,CAAC;4BAC1D,oBAAoB,GAAG,cAAc,CAAC,cAAc,CAAC,MAAM,CAAC,CAAC;4BAC7D,MAAM;wBACV,KAAK,uCAAkB,CAAC,IAAI;4BACxB,IAAI,GAAG,GAAG,IAAI,eAAM,CAAC,MAAM,CAAC,IAAI,EAAE,EAAE,GAAG,CAAC,CAAC;4BACzC,MAAM,CAAC,IAAI,EAAE,CAAC;4BACd,oBAAoB,GAAG,GAAG,CAAC,UAAU,CAAC,MAAM,CAAC,cAAc,CAAC,MAAM,CAAC,CAAC,CAAC;4BACrE,MAAM;wBACV,KAAK,uCAAkB,CAAC,IAAI;4BACxB,MAAM,CAAC,cAAc,EAAE,CAAC;4BACxB,oBAAoB,GAAG,MAAM,CAAC;4BAC9B,MAAM;qBACb;oBACD,IAAI,SAAS,GAAG,IAAI,CAAC,cAAc,CAAC,oBAAoB,CAAC,CAAC;oBAC1D,IAAI,MAAM,GAAG,IAAI,eAAM,CAAC,SAAS,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC;oBAC3C,MAAM,CAAC,oBAAoB,CAAC,YAAY,CAAC,CAAC;oBAC1C,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;oBACnB,MAAM,CAAC,oBAAoB,CAAC,YAAY,CAAC,CAAC;oBAC1C,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;iBACtB;gBACD,IAAI,gCAAgC,GAAG,IAAI,CAAC,cAAc,CAAC,aAAa,CAAC,CAAC;gBAC1E,IAAI,gCAAgC,CAAC,WAAW,EAAE,GAAG,6BAA6B,CAAC,WAAW,EAAE,EAAE;oBAC9F,6BAA6B,GAAG,gCAAgC,CAAC;oBACjE,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;oBACvB,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;iBAC1B;gBACD,YAAY,IAAI,UAAU,CAAC,cAAc,EAAE,CAAC;aAC/C;YACD,IAAI,CAAC,CAAC,GAAG,KAAK,CAAC;YACf,IAAI,CAAC,CAAC,GAAG,KAAK,CAAC;QACnB,CAAC;QAED;;;WAGG;QACH,SAAS,CAAC,QAAgB;YACtB,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,CAAA;QAC/B,CAAC;KAEJ;IAjHD,8DAiHC"}
+{"version":3,"file":"MultiLayerPerceptronModel.js","sourceRoot":"","sources":["../../../source/Model/NeuralNetwork/MultiLayerPerceptronModel.ts"],"names":[],"mappings":";;;AAAA,mEAA8D;AAC9D,wDAAmD;AACnD,2EAAsE;AAGtE,2FAAsF;AACtF,wDAAmD;AACnD,wDAAmD;AACnD,oEAA+D;AAE/D,4DAAuD;AAEvD,MAAa,yBAA0B,SAAQ,6CAAqB;IAExD,CAAC,CAAQ;IACP,kBAAkB,CAAoB;IAEhD;;;;;OAKG;IACK,eAAe,CAAC,CAAS,EAAE,MAAc;QAC7C,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,oBAAoB,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC;QAC1D,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,oBAAoB,CAAC,IAAI,CAAC,CAAC,EAAE,CAAC,GAAG,CAAC,EAAE,MAAM,CAAC,CAAC;IAC9D,CAAC;IAED;;;OAGG;IACH,YAAY,CAAC,QAAgB;QACzB,IAAI,KAAK,GAAG,IAAI,2BAAY,CAAC,QAAQ,CAAC,CAAA;QACtC,IAAI,CAAC,eAAe,CAAC,KAAK,CAAC,CAAA;QAC3B,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;QAC/B,IAAI,CAAC,CAAC,GAAG,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;QAC/B,IAAI,CAAC,kBAAkB,GAAG,IAAI,CAAC,sBAAsB,CAAC,KAAK,CAAC,CAAA;IAChE,CAAC;IAED;;OAEG;IACO,eAAe;QACrB,IAAI,CAAC,iCAAiC,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,kBAAkB,CAAC,CAAC;IACpF,CAAC;IAED,OAAO,CAAC,QAAgB;IACxB,CAAC;IAED;;;;;;;OAOG;IACH,KAAK,CAAC,KAAmB,EAAE,MAAiB;QACxC,IAAI,CAAC,UAAU,CAAC,KAAK,CAAC,CAAA;QACtB,IAAI,UAAU,GAAoC,MAAO,CAAA;QACzD,IAAI,SAAS,GAAG,IAAI,qBAAS,CAAC,KAAK,EAAE,UAAU,CAAC,uBAAuB,EAAE,EAAE,IAAI,CAAC,CAAC;QACjF,IAAI,QAAQ,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;QAC/B,IAAI,aAAa,GAAG,SAAS,CAAC,GAAG,CAAC,CAAC,CAAC,CAAA;QACpC,IAAI,CAAC,kBAAkB,GAAG,UAAU,CAAC,qBAAqB,EAAE,CAAC;QAC7D,IAAI,CAAC,eAAe,CAAC,UAAU,CAAC,cAAc,EAAE,EAAE,IAAI,eAAM,CAAC,UAAU,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;QACpF,IAAI,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;QAC3B,IAAI,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;QAC3B,IAAI,6BAA6B,GAAG,IAAI,qDAAyB,CAAC,GAAG,CAAC,CAAC;QACvE,IAAI,KAAK,GAAG,UAAU,CAAC,QAAQ,EAAE,CAAC;QAClC,IAAI,YAAY,GAAG,UAAU,CAAC,eAAe,EAAE,CAAC;QAChD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,KAAK,EAAE,CAAC,EAAE,EAAE,CAAC;YAC7B,QAAQ,CAAC,OAAO,CAAC,IAAI,eAAM,CAAC,UAAU,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC;YACnD,KAAK,IAAI,CAAC,GAAG,CAAC,EAAE,CAAC,GAAG,QAAQ,CAAC,IAAI,EAAE,EAAE,CAAC,EAAE,EAAE,CAAC;gBACvC,IAAI,CAAC,iBAAiB,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,CAAC,CAAC;gBACxC,IAAI,MAAM,GAAG,IAAI,CAAC,eAAe,CAAC,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,CAAC,EAAE,IAAI,CAAC,kBAAkB,CAAC,CAAC;gBAC3E,IAAI,YAAY,GAAG,MAAM,CAAC,MAAM,EAAE,CAAC;gBACnC,IAAI,OAAO,GAAG,IAAI,CAAC,gBAAgB,CAAC,QAAQ,CAAC,GAAG,CAAC,CAAC,CAAC,EAAE,YAAY,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC;gBAC3E,IAAI,MAAM,GAAG,IAAI,eAAM,CAAC,OAAO,EAAE,YAAY,CAAC,CAAC;gBAC/C,IAAI,IAAI,GAAG,IAAI,CAAC,CAAC,CAAC,0BAA0B,CAAC,OAAO,CAAC,CAAC;gBACtD,IAAI,CAAC,MAAM,CAAC,CAAC,CAAC,CAAC;gBACf,IAAI,oBAAoB,CAAA;gBACxB,QAAQ,IAAI,CAAC,kBAAkB,EAAC,CAAC;oBAC7B,KAAK,uCAAkB,CAAC,OAAO,CAAC;oBAChC;wBACI,IAAI,cAAc,GAAG,IAAI,CAAC,uBAAuB,CAAC,MAAM,CAAC,CAAC;wBAC1D,oBAAoB,GAAG,cAAc,CAAC,cAAc,CAAC,MAAM,CAAC,CAAC;wBAC7D,MAAM;oBACV,KAAK,uCAAkB,CAAC,IAAI;wBACxB,IAAI,GAAG,GAAG,IAAI,eAAM,CAAC,MAAM,CAAC,IAAI,EAAE,EAAE,GAAG,CAAC,CAAC;wBACzC,MAAM,CAAC,IAAI,EAAE,CAAC;wBACd,oBAAoB,GAAG,GAAG,CAAC,UAAU,CAAC,MAAM,CAAC,cAAc,CAAC,MAAM,CAAC,CAAC,CAAC;wBACrE,MAAM;oBACV,KAAK,uCAAkB,CAAC,IAAI;wBACxB,MAAM,CAAC,cAAc,EAAE,CAAC;wBACxB,oBAAoB,GAAG,MAAM,CAAC;wBAC9B,MAAM;gBACd,CAAC;gBACD,IAAI,SAAS,GAAG,IAAI,CAAC,cAAc,CAAC,oBAAoB,CAAC,CAAC;gBAC1D,IAAI,MAAM,GAAG,IAAI,eAAM,CAAC,SAAS,EAAE,IAAI,CAAC,CAAC,CAAC,CAAC;gBAC3C,MAAM,CAAC,oBAAoB,CAAC,YAAY,CAAC,CAAC;gBAC1C,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;gBACnB,MAAM,CAAC,oBAAoB,CAAC,YAAY,CAAC,CAAC;gBAC1C,IAAI,CAAC,CAAC,CAAC,GAAG,CAAC,MAAM,CAAC,CAAC;YACvB,CAAC;YACD,IAAI,gCAAgC,GAAG,IAAI,CAAC,cAAc,CAAC,aAAa,CAAC,CAAC;YAC1E,IAAI,gCAAgC,CAAC,WAAW,EAAE,GAAG,6BAA6B,CAAC,WAAW,EAAE,EAAE,CAAC;gBAC/F,6BAA6B,GAAG,gCAAgC,CAAC;gBACjE,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;gBACvB,KAAK,GAAG,IAAI,CAAC,CAAC,CAAC,KAAK,EAAE,CAAC;YAC3B,CAAC;YACD,YAAY,IAAI,UAAU,CAAC,cAAc,EAAE,CAAC;QAChD,CAAC;QACD,IAAI,CAAC,CAAC,GAAG,KAAK,CAAC;QACf,IAAI,CAAC,CAAC,GAAG,KAAK,CAAC;IACnB,CAAC;IAED;;;OAGG;IACH,SAAS,CAAC,QAAgB;QACtB,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,CAAA;IAC/B,CAAC;CAEJ;AAjHD,8DAiHC"}