/**
 * Copyright (c) 2011, The University of Southampton and the individual contributors.
 * All rights reserved.
 *
 * Redistribution and use in source and binary forms, with or without modification,
 * are permitted provided that the following conditions are met:
 *
 *   *  Redistributions of source code must retain the above copyright notice,
 *      this list of conditions and the following disclaimer.
 *
 *   *  Redistributions in binary form must reproduce the above copyright notice,
 *      this list of conditions and the following disclaimer in the documentation
 *      and/or other materials provided with the distribution.
 *
 *   *  Neither the name of the University of Southampton nor the names of its
 *      contributors may be used to endorse or promote products derived from this
 *      software without specific prior written permission.
 *
 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON
 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 */
package org.openimaj.ml.timeseries.processor;

import org.openimaj.ml.timeseries.processor.interpolation.LinearInterpolationProcessor;
import org.openimaj.ml.timeseries.series.DoubleTimeSeries;

/**
 * Calculates a moving average over a specified window in the past such that  
 * 
 * data[t_n] = sum^{m}_{i=1}{data[t_{n-i}}
 * 
 * This processor returns a value for each time in the underlying time series. 
 * For sensible results, consider interpolating a consistent time span using an {@link LinearInterpolationProcessor}
 * followed by this processor.
 * 
 * @author Sina Samangooei (ss@ecs.soton.ac.uk)
 *
 */
public class GaussianTimeSeriesProcessor implements TimeSeriesProcessor<double[],Double,DoubleTimeSeries>
{
        private double[] kernel;
        /**
         * The default number of sigmas at which the Gaussian function is truncated
         * when building a kernel
         */
        public static final double DEFAULT_GAUSS_TRUNCATE = 4.0d;
        /**
         * @param sigma the sigma of the guassian function
         */
        public GaussianTimeSeriesProcessor(double sigma) {
                this.kernel = makeKernel(sigma,DEFAULT_GAUSS_TRUNCATE);
        }
        
        /**
         * Construct a zero-mean Gaussian with the specified standard deviation.
         * @param sigma the standard deviation of the Gaussian
         * @param truncate the number of sigmas from the centre at which to
         *                              truncate the Gaussian 
         * @return an array representing a Gaussian function
         */
        public static double[] makeKernel(double sigma, double truncate) {
                if(sigma == 0) return new double[]{1f};
                //The kernel is truncated at truncate sigmas from center.
                int ksize = (int) (2.0f * truncate * sigma + 1.0f);
//              ksize = Math.max(1, ksize); // size must be at least 3
                if( ksize % 2 == 0 ) ksize++;  // size must be odd

                double [] kernel = new double[ksize];

                //build kernel
                float sum = 0.0f;
                for(int i = 0; i < ksize; i++) {
                        float x = i - ksize / 2;
                        kernel[i] = (float) Math.exp( -x * x / (2.0 * sigma * sigma) );
                        sum += kernel[i];
                }

                //normalise area to 1
                for(int i = 0; i < ksize; i++) {
                        kernel[i] /= sum;
                }
                
                return kernel;
        }
        
        /**
         * Convolve a double array
         * 
         * @param data the image to convolve.
         * @param kernel the convolution kernel.
         */
        public static void convolveHorizontal(double[] data, double [] kernel) {
                int halfsize = kernel.length / 2;

                double buffer[] = new double[data.length + kernel.length];              
                
                for(int i = 0; i < halfsize; i++)
                        buffer[i] = data[0];
                for(int i = 0; i < data.length; i++)
                        buffer[halfsize + i] = data[i];
                
                for(int i = 0; i < halfsize; i++)
                        buffer[halfsize + data.length + i] = data[data.length- 1];

//              convolveBuffer(buffer, kernel);
                int l =  buffer.length-kernel.length;
                for(int i = 0; i < l; i++) {
                        float sum = 0.0f;

                        for(int j = 0, jj=kernel.length-1; j < kernel.length; j++, jj--)
                                sum += buffer[i + j] * kernel[jj];

                        buffer[i] = sum;
                }
//              end convolveBuffer(buffer, kernel);

                for(int c=0; c<data.length; c++) data[c] = buffer[c];
        }

        @Override
        public void process(DoubleTimeSeries series) {
                convolveHorizontal(series.getData(),this.kernel);
        }
}