C# LinearAlgebra.Vector类代码示例

        /// <summary>Generate Logistic Regression model based on a set of examples.</summary>
        /// <param name="x">The Matrix to process.</param>
        /// <param name="y">The Vector to process.</param>
        /// <returns>Model.</returns>
        public override IModel Generate(Matrix x, Vector y)
        {
            // create initial theta
            Matrix copy = x.Copy();

            copy = PreProcessing.FeatureDimensions.IncreaseDimensions(copy, PolynomialFeatures);

            // add intercept term
            copy = copy.Insert(Vector.Ones(copy.Rows), 0, VectorType.Col);

            Vector theta = Vector.Ones(copy.Cols);

            var run = numl.Math.Optimization.GradientDescent.Run(theta, copy, y, this.MaxIterations, this.LearningRate, new numl.Math.Functions.Cost.LogisticCostFunction(), 
                this.Lambda, new numl.Math.Functions.Regularization.Regularization());

            LogisticRegressionModel model = new LogisticRegressionModel()
            {
                Descriptor = this.Descriptor,
                Theta = run.Item2,
                LogisticFunction = new Math.Functions.Logistic(),
                PolynomialFeatures = this.PolynomialFeatures
            };

            return model;
        }

        /// <summary>Predicts the given o.</summary>
        /// <exception cref="InvalidOperationException">Thrown when the requested operation is invalid.</exception>
        /// <param name="y">The Vector to process.</param>
        /// <returns>An object.</returns>
        public override double Predict(Vector y)
        {
            if (this.Root == null || this.Descriptor == null)
            {
                throw new InvalidOperationException("Invalid Model - Missing information");
            }

            var lp = Vector.Zeros(this.Root.Probabilities.Length);
            for (var i = 0; i < this.Root.Probabilities.Length; i++)
            {
                var stat = this.Root.Probabilities[i];
                lp[i] = Math.Log(stat.Probability);
                for (var j = 0; j < y.Length; j++)
                {
                    var conditional = stat.Conditionals[j];
                    var p = conditional.GetStatisticFor(y[j]);

                    // check for missing range, assign bad probability
                    lp[i] += Math.Log(p == null ? 10e-10 : p.Probability);
                }
            }

            var idx = lp.MaxIndex();
            return this.Root.Probabilities[idx].X.Min;
        }

        /// <summary>Generate model based on a set of examples.</summary>
        /// <param name="x">The Matrix to process.</param>
        /// <param name="y">The Vector to process.</param>
        /// <returns>Model.</returns>
        public override IModel Generate(Matrix x, Vector y)
        {
            // because I said so...
            if (this.MaxIterations == -1)
            {
                this.MaxIterations = x.Rows * 1000;
            }

            var network = Network.Default(this.Descriptor, x, y, this.Activation);
            var model = new NeuralNetworkModel { Descriptor = this.Descriptor, Network = network };
            this.OnModelChanged(this, ModelEventArgs.Make(model, "Initialized"));

            for (var i = 0; i < this.MaxIterations; i++)
            {
                var idx = i % x.Rows;
                network.Forward(x[idx, VectorType.Row]);

                // OnModelChanged(this, ModelEventArgs.Make(model, "Forward"));
                network.Back(y[idx], this.LearningRate);
                var output = string.Format("Run ({0}/{1})", i, this.MaxIterations);
                this.OnModelChanged(this, ModelEventArgs.Make(model, output));
            }

            return model;
        }

        /// <summary>
        ///     Performs gradient descent to optomise theta parameters.
        /// </summary>
        /// <param name="theta">Initial Theta (Zeros)</param>
        /// <param name="x">Training set</param>
        /// <param name="y">Training labels</param>
        /// <param name="maxIterations">Maximum number of iterations to run gradient descent</param>
        /// <param name="learningRateAlpha">The learning rate (Alpha)</param>
        /// <param name="costFunction">Cost function to use for gradient descent</param>
        /// <param name="lambda">The regularization constant to apply</param>
        /// <param name="regularizer">The regularization function to apply</param>
        /// <returns></returns>
        public static Tuple<double, Vector> Run(
            Vector theta, 
            Matrix x, 
            Vector y, 
            int maxIterations, 
            double learningRateAlpha, 
            ICostFunction costFunction, 
            double lambda, 
            IRegularizer regularizer)
        {
            var bestTheta = theta.Copy();
            var bestCost = double.PositiveInfinity;

            double currentCost = 0;
            var currentGradient = theta.Copy();

            for (var i = 0; i <= maxIterations; i++)
            {
                currentCost = costFunction.ComputeCost(bestTheta, x, y, lambda, regularizer);
                currentGradient = costFunction.ComputeGradient(bestTheta, x, y, lambda, regularizer);

                if (currentCost < bestCost)
                {
                    bestTheta = bestTheta - learningRateAlpha * currentGradient;
                    bestCost = currentCost;
                }
                else
                {
                    learningRateAlpha = learningRateAlpha * 0.99;
                }
            }

            return new Tuple<double, Vector>(bestCost, bestTheta);
        }

        //[Fact]
        public void Test_SVM_Email_Classification()
        {
            var training_Data = System.IO.File.ReadAllLines("Data\\Emails\\Training_Data.txt")
                                    .Select(s => s.Split(new char[] { ' ' }, StringSplitOptions.RemoveEmptyEntries).ToVector(f => double.Parse(f.Trim()))).ToMatrix();

            var training_Labels = System.IO.File.ReadAllLines("Data\\Emails\\Training_Labels.txt")
                                    .Select(s => double.Parse(s.Trim())).ToVector();

            var test_Data = System.IO.File.ReadAllLines("Data\\Emails\\Test_Data.txt")
                                    .Select(s => s.Split(new char[] { ' ' }, StringSplitOptions.RemoveEmptyEntries).ToVector(f => double.Parse(f.Trim()))).ToMatrix();

            var test_Labels = System.IO.File.ReadAllLines("Data\\Emails\\Test_Labels.txt")
                                    .Select(s => double.Parse(s.Trim())).ToVector();

            var generator = new Supervised.SVM.SVMGenerator()
            {
                C = 0.1,
                MaxIterations = 5,
                SelectionFunction = new Supervised.SVM.Selection.RandomSetSelection()
            };

            var model = generator.Generate(training_Data, training_Labels);

            Vector predictions = new Vector(test_Labels.Count());
            for (int x = 0; x < test_Labels.Count(); x++)
            {
                predictions[x] = model.Predict(test_Data[x]);
            }

            var score = numl.Supervised.Score.ScorePredictions(predictions, test_Labels);

            Console.WriteLine($"SVM Model\n: { score }");
        }

 /// <summary>Products the given v.</summary>
 /// <param name="v">A variable-length parameters list containing v.</param>
 /// <returns>A double.</returns>
 public static double Prod(Vector v)
 {
     var prod = v[0];
     for (int i = 1; i < v.Length; i++)
         prod *= v[i];
     return prod;
 }

 /// <summary>Sums the given v.</summary>
 /// <param name="v">A variable-length parameters list containing v.</param>
 /// <returns>A double.</returns>
 public static double Sum(Vector v)
 {
     double sum = 0;
     for (int i = 0; i < v.Length; i++)
         sum += v[i];
     return sum;
 }

        public void Logistic_Regression_Test_CostFunction_1()
        {
            Matrix X = new[,]
            {{ 1, 1, 1 },
             { 1, 1, 1 },
             { 1, 1, 1 },
             { 8, 1, 6 },
             { 3, 5 ,7 },
             { 4, 9, 2 }};

            Vector y = new Vector(new double[] { 1, 0, 1, 0, 1, 0 });
            Vector theta = new Vector(new double[] { 0, 1, 0 });

            ICostFunction logisticCostFunction = new LogisticCostFunction()
            {
                X = X,
                Y = y,
                Lambda = 3,
                Regularizer = new L2Regularizer()
            };

            double cost = logisticCostFunction.ComputeCost(theta.Copy());

            theta = logisticCostFunction.ComputeGradient(theta.Copy());

            Assert.Equal(2.2933d, System.Math.Round(cost, 4));

            Assert.Equal(1.6702d, System.Math.Round(theta[0], 4));
            Assert.Equal(2.1483d, System.Math.Round(theta[1], 4));
            Assert.Equal(1.0887d, System.Math.Round(theta[2], 4));
        }

 public static Matrix Diag(Vector v)
 {
     Matrix m = Matrix.Zeros(v.Length);
     for (int i = 0; i < v.Length; i++)
         m[i, i] = v[i];
     return m;
 }

 public static bool ContainsNaN(Vector vector)
 {
     for (int i = 0; i < vector.Length; i++)
         if(double.IsNaN(vector[i]))
             return true;
     return false;
 }

        /// <summary>
        /// Calculates segmented conditional impurity of y | x When stipulating segments (s), X is broken
        /// up into s many segments therefore P(X=x_s) becomes a range probability rather than a fixed
        /// probability. In essence the average over H(Y|X = x) becomes SUM_s [ p_s * H(Y|X = x_s) ]. The
        /// values that were used to do the split are stored in the Splits member.
        /// </summary>
        /// <exception cref="InvalidOperationException">Thrown when the requested operation is invalid.</exception>
        /// <param name="y">Target impurity.</param>
        /// <param name="x">Conditioned impurity.</param>
        /// <param name="segments">Number of segments over x to condition upon.</param>
        /// <returns>Segmented conditional impurity measure.</returns>
        public double SegmentedConditional(Vector y, Vector x, int segments)
        {
            if (x == null && y == null)
                throw new InvalidOperationException("x and y do not exist!");

            return SegmentedConditional(y, x, x.Segment(segments));
        }

        /// <summary>
        /// Calculates conditional impurity of y | x
        /// R(Y|X) is the average of H(Y|X = x) over all possible values
        /// X may take. 
        /// </summary>
        /// <param name="y">Target impurity</param>
        /// <param name="x">Conditioned impurity</param>
        /// <param name="width">Split of values over x to condition upon</param>
        /// <returns>Conditional impurity measure</returns>
        public double Conditional(Vector y, Vector x)
        {
            if (x == null && y == null)
                throw new InvalidOperationException("x and y do not exist!");

            double p = 0,               // probability of slice
                   h = 0,               // impurity of y | x_i : ith slice
                   result = 0,          // aggregated sum
                   count = x.Count();   // total items in list

            var values = x.Distinct().OrderBy(z => z);  // distinct values to split on

            Segments = values.Select(z => Range.Make(z, z)).ToArray();
            Discrete = true;

            // for each distinct value
            // calculate conditional impurity
            // and aggregate results
            foreach (var i in values)
            {
                // get slice
                var s = x.Indices(d => d == i);
                // slice probability
                p = (double)s.Count() / (double)count;
                // impurity of (y | x_i)
                h = Calculate(y.Slice(s));
                // sum up
                result += p * h;
            }

            return result;
        }

 public void Test_Vector_Slicing_With_Indices(IEnumerable<double> source, IEnumerable<int> indices, IEnumerable<double> truth)
 {
     var x = new Vector(source);
     var t = new Vector(truth);
     var slice = x.Slice(indices);
     Assert.AreEqual(t, slice);
 }

        /// <summary>Generate Linear Regression model based on a set of examples.</summary>
        /// <param name="x">The Matrix to process.</param>
        /// <param name="y">The Vector to process.</param>
        /// <returns>Model.</returns>
        public override IModel Generate(Matrix x, Vector y)
        {
            // create initial theta
            Vector theta = Vector.Ones(x.Cols + 1);
            Matrix copy = x.Copy();

            // normalise features
            for (int i = 0; i < copy.Cols; i++)
            {
                var j = FeatureNormalizer.FeatureScale(copy[i, VectorType.Col]);
                for (int k = 0; k < copy.Rows; k++)
                {
                    copy[k, i] = j[k];
                }
            }

            // add intercept term
            copy = copy.Insert(Vector.Ones(copy.Rows), 0, VectorType.Col);

            // run gradient descent
            var run = GradientDescent.Run(theta, copy, y, MaxIterations, LearningRate, new LinearCostFunction(),
                Lambda, new Regularization());

            // once converged create model and apply theta

            LinearRegressionModel model = new LinearRegressionModel(x.Mean(VectorType.Row), x.StdDev(VectorType.Row))
            {
                Descriptor = Descriptor,
                Theta = run.Item2
            };

            return model;
        }

        /// <summary>
        /// Initializes a new Optimizer using the default values.
        /// <param name="theta">Theta to optimize.</param>
        /// <param name="maxIterations">Maximum number of iterations.</param>
        /// <param name="learningRate">Learning Rate (alpha) (Optional).</param>
        /// <param name="momentum">Momentum parameter for use in accelerated methods (Optional).</param>
        /// <param name="optimizationMethod">Type of optimization method to use (Optional).</param>
        /// <param name="optimizer">An external typed optimization method to use (Optional).</param>
        /// </summary>
        public Optimizer(Vector theta, int maxIterations, double learningRate = 1.0, double momentum = 0.9,
            OptimizationMethods optimizationMethod = OptimizationMethods.StochasticGradientDescent, OptimizationMethod optimizer = null)
        {
            this.Completed = false;
            if (optimizationMethod != OptimizationMethods.External)
            {
                switch (optimizationMethod)
                {
                    case OptimizationMethods.FastGradientDescent: optimizer = new numl.Math.Optimization.Methods.GradientDescent.FastGradientDescent() { Momentum = momentum }; break;
                    case OptimizationMethods.StochasticGradientDescent: optimizer = new numl.Math.Optimization.Methods.GradientDescent.StochasticGradientDescent(); break;
                    case OptimizationMethods.NAGDescent: optimizer = new numl.Math.Optimization.Methods.GradientDescent.NAGDescent() { Momentum = momentum }; break;
                }
            }

            this.OpimizationMethod = optimizer;

            this.Properties = new OptimizerProperties()
            {
                Iteration = 0,
                MaxIterations = maxIterations,
                Cost = double.MaxValue,
                Gradient = Vector.Zeros(theta.Length),
                Theta = theta,
                LearningRate = learningRate,
                Momentum = momentum
            };
        }