FastFourierTransformer.java

/*
 * Licensed to the Apache Software Foundation (ASF) under one or more
 * contributor license agreements.  See the NOTICE file distributed with
 * this work for additional information regarding copyright ownership.
 * The ASF licenses this file to You under the Apache License, Version 2.0
 * (the "License"); you may not use this file except in compliance with
 * the License.  You may obtain a copy of the License at
 *
 *      https://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

/*
 * This is not the original file distributed by the Apache Software Foundation
 * It has been modified by the Hipparchus project
 */
package org.hipparchus.transform;

import java.io.Serializable;

import org.hipparchus.analysis.FunctionUtils;
import org.hipparchus.analysis.UnivariateFunction;
import org.hipparchus.complex.Complex;
import org.hipparchus.exception.MathIllegalArgumentException;
import org.hipparchus.exception.MathRuntimeException;
import org.hipparchus.util.ArithmeticUtils;
import org.hipparchus.util.FastMath;
import org.hipparchus.util.MathArrays;
import org.hipparchus.util.MathUtils;

/**
 * Implements the Fast Fourier Transform for transformation of one-dimensional
 * real or complex data sets. For reference, see <em>Applied Numerical Linear
 * Algebra</em>, ISBN 0898713897, chapter 6.
 * <p>
 * There are several variants of the discrete Fourier transform, with various
 * normalization conventions, which are specified by the parameter
 * {@link DftNormalization}.
 * <p>
 * The current implementation of the discrete Fourier transform as a fast
 * Fourier transform requires the length of the data set to be a power of 2.
 * This greatly simplifies and speeds up the code. Users can pad the data with
 * zeros to meet this requirement. There are other flavors of FFT, for
 * reference, see S. Winograd,
 * <i>On computing the discrete Fourier transform</i>, Mathematics of
 * Computation, 32 (1978), 175 - 199.
 *
 * @see DftNormalization
 */
public class FastFourierTransformer implements Serializable {

    /** Serializable version identifier. */
    private static final long serialVersionUID = 20120210L;

    /**
     * {@code W_SUB_N_R[i]} is the real part of
     * {@code exp(- 2 * i * pi / n)}:
     * {@code W_SUB_N_R[i] = cos(2 * pi/ n)}, where {@code n = 2^i}.
     */
    private static final double[] W_SUB_N_R =
            {  0x1.0p0, -0x1.0p0, 0x1.1a62633145c07p-54, 0x1.6a09e667f3bcdp-1
            , 0x1.d906bcf328d46p-1, 0x1.f6297cff75cbp-1, 0x1.fd88da3d12526p-1, 0x1.ff621e3796d7ep-1
            , 0x1.ffd886084cd0dp-1, 0x1.fff62169b92dbp-1, 0x1.fffd8858e8a92p-1, 0x1.ffff621621d02p-1
            , 0x1.ffffd88586ee6p-1, 0x1.fffff62161a34p-1, 0x1.fffffd8858675p-1, 0x1.ffffff621619cp-1
            , 0x1.ffffffd885867p-1, 0x1.fffffff62161ap-1, 0x1.fffffffd88586p-1, 0x1.ffffffff62162p-1
            , 0x1.ffffffffd8858p-1, 0x1.fffffffff6216p-1, 0x1.fffffffffd886p-1, 0x1.ffffffffff621p-1
            , 0x1.ffffffffffd88p-1, 0x1.fffffffffff62p-1, 0x1.fffffffffffd9p-1, 0x1.ffffffffffff6p-1
            , 0x1.ffffffffffffep-1, 0x1.fffffffffffffp-1, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0, 0x1.0p0
            , 0x1.0p0, 0x1.0p0, 0x1.0p0 };

    /**
     * {@code W_SUB_N_I[i]} is the imaginary part of
     * {@code exp(- 2 * i * pi / n)}:
     * {@code W_SUB_N_I[i] = -sin(2 * pi/ n)}, where {@code n = 2^i}.
     */
    private static final double[] W_SUB_N_I =
            {  0x1.1a62633145c07p-52, -0x1.1a62633145c07p-53, -0x1.0p0, -0x1.6a09e667f3bccp-1
            , -0x1.87de2a6aea963p-2, -0x1.8f8b83c69a60ap-3, -0x1.917a6bc29b42cp-4, -0x1.91f65f10dd814p-5
            , -0x1.92155f7a3667ep-6, -0x1.921d1fcdec784p-7, -0x1.921f0fe670071p-8, -0x1.921f8becca4bap-9
            , -0x1.921faaee6472dp-10, -0x1.921fb2aecb36p-11, -0x1.921fb49ee4ea6p-12, -0x1.921fb51aeb57bp-13
            , -0x1.921fb539ecf31p-14, -0x1.921fb541ad59ep-15, -0x1.921fb5439d73ap-16, -0x1.921fb544197ap-17
            , -0x1.921fb544387bap-18, -0x1.921fb544403c1p-19, -0x1.921fb544422c2p-20, -0x1.921fb54442a83p-21
            , -0x1.921fb54442c73p-22, -0x1.921fb54442cefp-23, -0x1.921fb54442d0ep-24, -0x1.921fb54442d15p-25
            , -0x1.921fb54442d17p-26, -0x1.921fb54442d18p-27, -0x1.921fb54442d18p-28, -0x1.921fb54442d18p-29
            , -0x1.921fb54442d18p-30, -0x1.921fb54442d18p-31, -0x1.921fb54442d18p-32, -0x1.921fb54442d18p-33
            , -0x1.921fb54442d18p-34, -0x1.921fb54442d18p-35, -0x1.921fb54442d18p-36, -0x1.921fb54442d18p-37
            , -0x1.921fb54442d18p-38, -0x1.921fb54442d18p-39, -0x1.921fb54442d18p-40, -0x1.921fb54442d18p-41
            , -0x1.921fb54442d18p-42, -0x1.921fb54442d18p-43, -0x1.921fb54442d18p-44, -0x1.921fb54442d18p-45
            , -0x1.921fb54442d18p-46, -0x1.921fb54442d18p-47, -0x1.921fb54442d18p-48, -0x1.921fb54442d18p-49
            , -0x1.921fb54442d18p-50, -0x1.921fb54442d18p-51, -0x1.921fb54442d18p-52, -0x1.921fb54442d18p-53
            , -0x1.921fb54442d18p-54, -0x1.921fb54442d18p-55, -0x1.921fb54442d18p-56, -0x1.921fb54442d18p-57
            , -0x1.921fb54442d18p-58, -0x1.921fb54442d18p-59, -0x1.921fb54442d18p-60 };

    /** The type of DFT to be performed. */
    private final DftNormalization normalization;

    /**
     * Creates a new instance of this class, with various normalization
     * conventions.
     *
     * @param normalization the type of normalization to be applied to the
     * transformed data
     */
    public FastFourierTransformer(final DftNormalization normalization) {
        this.normalization = normalization;
    }

    /**
     * Performs identical index bit reversal shuffles on two arrays of identical
     * size. Each element in the array is swapped with another element based on
     * the bit-reversal of the index. For example, in an array with length 16,
     * item at binary index 0011 (decimal 3) would be swapped with the item at
     * binary index 1100 (decimal 12).
     *
     * @param a the first array to be shuffled
     * @param b the second array to be shuffled
     */
    private static void bitReversalShuffle2(double[] a, double[] b) {
        final int n = a.length;
        assert b.length == n;
        final int halfOfN = n >> 1;

        int j = 0;
        for (int i = 0; i < n; i++) {
            if (i < j) {
                // swap indices i & j
                double temp = a[i];
                a[i] = a[j];
                a[j] = temp;

                temp = b[i];
                b[i] = b[j];
                b[j] = temp;
            }

            int k = halfOfN;
            while (k <= j && k > 0) {
                j -= k;
                k >>= 1;
            }
            j += k;
        }
    }

    /**
     * Applies the proper normalization to the specified transformed data.
     *
     * @param dataRI the unscaled transformed data
     * @param normalization the normalization to be applied
     * @param type the type of transform (forward, inverse) which resulted in the specified data
     */
    private static void normalizeTransformedData(final double[][] dataRI,
        final DftNormalization normalization, final TransformType type) {

        final double[] dataR = dataRI[0];
        final double[] dataI = dataRI[1];
        final int n = dataR.length;
        assert dataI.length == n;

        switch (normalization) {
            case STANDARD:
                if (type == TransformType.INVERSE) {
                    final double scaleFactor = 1.0 / n;
                    for (int i = 0; i < n; i++) {
                        dataR[i] *= scaleFactor;
                        dataI[i] *= scaleFactor;
                    }
                }
                break;
            case UNITARY:
                final double scaleFactor = 1.0 / FastMath.sqrt(n);
                for (int i = 0; i < n; i++) {
                    dataR[i] *= scaleFactor;
                    dataI[i] *= scaleFactor;
                }
                break;
            default:
                // This should never occur in normal conditions. However this
                // clause has been added as a safeguard if other types of
                // normalizations are ever implemented, and the corresponding
                // test is forgotten in the present switch.
                throw MathRuntimeException.createInternalError();
        }
    }

    /**
     * Computes the standard transform of the specified complex data. The
     * computation is done in place. The input data is laid out as follows
     * <ul>
     *   <li>{@code dataRI[0][i]} is the real part of the {@code i}-th data point,</li>
     *   <li>{@code dataRI[1][i]} is the imaginary part of the {@code i}-th data point.</li>
     * </ul>
     *
     * @param dataRI the two dimensional array of real and imaginary parts of the data
     * @param normalization the normalization to be applied to the transformed data
     * @param type the type of transform (forward, inverse) to be performed
     * @throws MathIllegalArgumentException if the number of rows of the specified
     *   array is not two, or the array is not rectangular
     * @throws MathIllegalArgumentException if the number of data points is not
     *   a power of two
     */
    public static void transformInPlace(final double[][] dataRI,
        final DftNormalization normalization, final TransformType type) {

        MathUtils.checkDimension(dataRI.length, 2);
        final double[] dataR = dataRI[0];
        final double[] dataI = dataRI[1];
        MathArrays.checkEqualLength(dataR, dataI);

        final int n = dataR.length;
        if (!ArithmeticUtils.isPowerOfTwo(n)) {
            throw new MathIllegalArgumentException(LocalizedFFTFormats.NOT_POWER_OF_TWO_CONSIDER_PADDING,
                                                   Integer.valueOf(n));
        }

        if (n == 1) {
            return;
        } else if (n == 2) {
            final double srcR0 = dataR[0];
            final double srcI0 = dataI[0];
            final double srcR1 = dataR[1];
            final double srcI1 = dataI[1];

            // X_0 = x_0 + x_1
            dataR[0] = srcR0 + srcR1;
            dataI[0] = srcI0 + srcI1;
            // X_1 = x_0 - x_1
            dataR[1] = srcR0 - srcR1;
            dataI[1] = srcI0 - srcI1;

            normalizeTransformedData(dataRI, normalization, type);
            return;
        }

        bitReversalShuffle2(dataR, dataI);

        // Do 4-term DFT.
        if (type == TransformType.INVERSE) {
            for (int i0 = 0; i0 < n; i0 += 4) {
                final int i1 = i0 + 1;
                final int i2 = i0 + 2;
                final int i3 = i0 + 3;

                final double srcR0 = dataR[i0];
                final double srcI0 = dataI[i0];
                final double srcR1 = dataR[i2];
                final double srcI1 = dataI[i2];
                final double srcR2 = dataR[i1];
                final double srcI2 = dataI[i1];
                final double srcR3 = dataR[i3];
                final double srcI3 = dataI[i3];

                // 4-term DFT
                // X_0 = x_0 + x_1 + x_2 + x_3
                dataR[i0] = srcR0 + srcR1 + srcR2 + srcR3;
                dataI[i0] = srcI0 + srcI1 + srcI2 + srcI3;
                // X_1 = x_0 - x_2 + j * (x_3 - x_1)
                dataR[i1] = srcR0 - srcR2 + (srcI3 - srcI1);
                dataI[i1] = srcI0 - srcI2 + (srcR1 - srcR3);
                // X_2 = x_0 - x_1 + x_2 - x_3
                dataR[i2] = srcR0 - srcR1 + srcR2 - srcR3;
                dataI[i2] = srcI0 - srcI1 + srcI2 - srcI3;
                // X_3 = x_0 - x_2 + j * (x_1 - x_3)
                dataR[i3] = srcR0 - srcR2 + (srcI1 - srcI3);
                dataI[i3] = srcI0 - srcI2 + (srcR3 - srcR1);
            }
        } else {
            for (int i0 = 0; i0 < n; i0 += 4) {
                final int i1 = i0 + 1;
                final int i2 = i0 + 2;
                final int i3 = i0 + 3;

                final double srcR0 = dataR[i0];
                final double srcI0 = dataI[i0];
                final double srcR1 = dataR[i2];
                final double srcI1 = dataI[i2];
                final double srcR2 = dataR[i1];
                final double srcI2 = dataI[i1];
                final double srcR3 = dataR[i3];
                final double srcI3 = dataI[i3];

                // 4-term DFT
                // X_0 = x_0 + x_1 + x_2 + x_3
                dataR[i0] = srcR0 + srcR1 + srcR2 + srcR3;
                dataI[i0] = srcI0 + srcI1 + srcI2 + srcI3;
                // X_1 = x_0 - x_2 + j * (x_3 - x_1)
                dataR[i1] = srcR0 - srcR2 + (srcI1 - srcI3);
                dataI[i1] = srcI0 - srcI2 + (srcR3 - srcR1);
                // X_2 = x_0 - x_1 + x_2 - x_3
                dataR[i2] = srcR0 - srcR1 + srcR2 - srcR3;
                dataI[i2] = srcI0 - srcI1 + srcI2 - srcI3;
                // X_3 = x_0 - x_2 + j * (x_1 - x_3)
                dataR[i3] = srcR0 - srcR2 + (srcI3 - srcI1);
                dataI[i3] = srcI0 - srcI2 + (srcR1 - srcR3);
            }
        }

        int lastN0 = 4;
        int lastLogN0 = 2;
        while (lastN0 < n) {
            int n0 = lastN0 << 1;
            int logN0 = lastLogN0 + 1;
            double wSubN0R = W_SUB_N_R[logN0];
            double wSubN0I = W_SUB_N_I[logN0];
            if (type == TransformType.INVERSE) {
                wSubN0I = -wSubN0I;
            }

            // Combine even/odd transforms of size lastN0 into a transform of size N0 (lastN0 * 2).
            for (int destEvenStartIndex = 0; destEvenStartIndex < n; destEvenStartIndex += n0) {
                int destOddStartIndex = destEvenStartIndex + lastN0;

                double wSubN0ToRR = 1;
                double wSubN0ToRI = 0;

                for (int r = 0; r < lastN0; r++) {
                    double grR = dataR[destEvenStartIndex + r];
                    double grI = dataI[destEvenStartIndex + r];
                    double hrR = dataR[destOddStartIndex + r];
                    double hrI = dataI[destOddStartIndex + r];

                    // dest[destEvenStartIndex + r] = Gr + WsubN0ToR * Hr
                    dataR[destEvenStartIndex + r] = grR + wSubN0ToRR * hrR - wSubN0ToRI * hrI;
                    dataI[destEvenStartIndex + r] = grI + wSubN0ToRR * hrI + wSubN0ToRI * hrR;
                    // dest[destOddStartIndex + r] = Gr - WsubN0ToR * Hr
                    dataR[destOddStartIndex + r] = grR - (wSubN0ToRR * hrR - wSubN0ToRI * hrI);
                    dataI[destOddStartIndex + r] = grI - (wSubN0ToRR * hrI + wSubN0ToRI * hrR);

                    // WsubN0ToR *= WsubN0R
                    double nextWsubN0ToRR = wSubN0ToRR * wSubN0R - wSubN0ToRI * wSubN0I;
                    double nextWsubN0ToRI = wSubN0ToRR * wSubN0I + wSubN0ToRI * wSubN0R;
                    wSubN0ToRR = nextWsubN0ToRR;
                    wSubN0ToRI = nextWsubN0ToRI;
                }
            }

            lastN0 = n0;
            lastLogN0 = logN0;
        }

        normalizeTransformedData(dataRI, normalization, type);
    }

    /**
     * Returns the (forward, inverse) transform of the specified real data set.
     *
     * @param f the real data array to be transformed
     * @param type the type of transform (forward, inverse) to be performed
     * @return the complex transformed array
     * @throws MathIllegalArgumentException if the length of the data array is not a power of two
     */
    public Complex[] transform(final double[] f, final TransformType type) {
        final double[][] dataRI = { f.clone(), new double[f.length] };
        transformInPlace(dataRI, normalization, type);
        return TransformUtils.createComplexArray(dataRI);
    }

    /**
     * Returns the (forward, inverse) transform of the specified real function,
     * sampled on the specified interval.
     *
     * @param f the function to be sampled and transformed
     * @param min the (inclusive) lower bound for the interval
     * @param max the (exclusive) upper bound for the interval
     * @param n the number of sample points
     * @param type the type of transform (forward, inverse) to be performed
     * @return the complex transformed array
     * @throws org.hipparchus.exception.MathIllegalArgumentException
     *   if the lower bound is greater than, or equal to the upper bound
     * @throws org.hipparchus.exception.MathIllegalArgumentException
     *   if the number of sample points {@code n} is negative
     * @throws MathIllegalArgumentException if the number of sample points
     *   {@code n} is not a power of two
     */
    public Complex[] transform(final UnivariateFunction f,
                               final double min, final double max, final int n,
                               final TransformType type) {

        final double[] data = FunctionUtils.sample(f, min, max, n);
        return transform(data, type);
    }

    /**
     * Returns the (forward, inverse) transform of the specified complex data set.
     *
     * @param f the complex data array to be transformed
     * @param type the type of transform (forward, inverse) to be performed
     * @return the complex transformed array
     * @throws MathIllegalArgumentException if the length of the data array is not a power of two
     */
    public Complex[] transform(final Complex[] f, final TransformType type) {
        final double[][] dataRI = TransformUtils.createRealImaginaryArray(f);

        transformInPlace(dataRI, normalization, type);

        return TransformUtils.createComplexArray(dataRI);
    }

}