LambdaMove

// ******************************************************************************
//
// Title:       Force Field X.
// Description: Force Field X - Software for Molecular Biophysics.
// Copyright:   Copyright (c) Michael J. Schnieders 2001-2024.
//
// This file is part of Force Field X.
//
// Force Field X is free software; you can redistribute it and/or modify it
// under the terms of the GNU General Public License version 3 as published by
// the Free Software Foundation.
//
// Force Field X is distributed in the hope that it will be useful, but WITHOUT
// ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS
// FOR A PARTICULAR PURPOSE. See the GNU General Public License for more
// details.
//
// You should have received a copy of the GNU General Public License along with
// Force Field X; if not, write to the Free Software Foundation, Inc., 59 Temple
// Place, Suite 330, Boston, MA 02111-1307 USA
//
// Linking this library statically or dynamically with other modules is making a
// combined work based on this library. Thus, the terms and conditions of the
// GNU General Public License cover the whole combination.
//
// As a special exception, the copyright holders of this library give you
// permission to link this library with independent modules to produce an
// executable, regardless of the license terms of these independent modules, and
// to copy and distribute the resulting executable under terms of your choice,
// provided that you also meet, for each linked independent module, the terms
// and conditions of the license of that module. An independent module is a
// module which is not derived from or based on this library. If you modify this
// library, you may extend this exception to your version of the library, but
// you are not obligated to do so. If you do not wish to do so, delete this
// exception statement from your version.
//
// ******************************************************************************
package ffx.algorithms.mc;

import static org.apache.commons.math3.util.FastMath.abs;
import static org.apache.commons.math3.util.FastMath.max;
import static org.apache.commons.math3.util.FastMath.min;

import ffx.algorithms.thermodynamics.OrthogonalSpaceTempering;
import java.util.Random;
import java.util.logging.Logger;

/**
 * Define an MC move to update lambda.
 *
 * @author Mallory R. Tollefson
 */
public class LambdaMove implements MCMove {

  private static final Logger logger = Logger.getLogger(LambdaMove.class.getName());
  /** Apply the Lambda move to an OST instance. */
  private final OrthogonalSpaceTempering orthogonalSpaceTempering;
  /** Random number generator. */
  private final Random random;
  /** Current value of lambda, which always refreshed from the OST instance. */
  private double currentLambda;
  /**
   * Lambda move size: 1) The standard deviation for continuous moves from a Gaussian distribution.
   * 2) The step size for discrete moves.
   */
  private double moveSize = 0.1;
  /** If true, do continuous moves. Otherwise, use discrete moves. */
  private boolean isContinuous = true;

  /**
   * Constructor for LambdaMove.
   *
   * @param orthogonalSpaceTempering a {@link OrthogonalSpaceTempering} object.
   */
  public LambdaMove(OrthogonalSpaceTempering orthogonalSpaceTempering) {
    this.orthogonalSpaceTempering = orthogonalSpaceTempering;
    random = new Random();
  }

  /**
   * Constructor for LambdaMove.
   *
   * @param randomSeed Random seed to use.
   * @param orthogonalSpaceTempering OrthogonalSpaceTempering instance.
   */
  public LambdaMove(int randomSeed, OrthogonalSpaceTempering orthogonalSpaceTempering) {
    this.orthogonalSpaceTempering = orthogonalSpaceTempering;
    random = new Random(randomSeed);
  }

  /**
   * Applies 0-1 mirroring conditions to lam + dL. Skips any moves where dL is greater than 1 or less
   * than -1, and skips 50% of moves from 0 or 1 (exact).
   *
   * @param random Source of randomness.
   * @param lam Initial lambda.
   * @param dL Change in lambda.
   * @return Correctly mirrored lam + dL
   */
  public static double mirror(Random random, double lam, double dL) {
    if (lam == 0.0 || lam == 1.0) {
      boolean skip = random.nextBoolean();
      if (skip) {
        return lam;
      }
    }
    // Eliminate really weird edge cases.
    if (abs(dL) > 1.0) {
      logger.warning(String.format(" Skipping large lambda move of %.3f not between -1 and +1", dL));
      return lam;
    }
    // Math.abs to mirror negative values.
    double newLam = abs(lam + dL);
    // If greater than 1, mirror via 2.0 - val
    return newLam <= 1.0 ? newLam : 2.0 - newLam;
  }

  /**
   * Get the Lambda move size, which is a standard deviation for continuous moves or step size for
   * discrete moves.
   *
   * @return The lambda move size.
   */
  public double getMoveSize() {
    return moveSize;
  }

  /**
   * Get the Lambda move size, which is a standard deviation for continuous moves or step size for
   * discrete moves.
   *
   * @param moveSize a double.
   */
  public void setMoveSize(double moveSize) {
    this.moveSize = moveSize;
  }

  /**
   * If true, do continuous moves. Otherwise, use discrete moves.
   *
   * @return Returns true if the lambda moves are continuous.
   */
  public boolean isContinuous() {
    return isContinuous;
  }

  /**
   * If true, do continuous moves. Otherwise, use discrete moves.
   *
   * @param continuous Sets the lambda move style.
   */
  public void setContinuous(boolean continuous) {
    isContinuous = continuous;
  }

  /** {@inheritDoc} */
  @Override
  public void move() {
    currentLambda = orthogonalSpaceTempering.getLambda();

    // Draw a trial move from the distribution.
    double dL = isContinuous ? continuousMove() : discreteMove();
    double newLambda = mirror(currentLambda, dL);

    // Update the OST instance.
    orthogonalSpaceTempering.setLambda(newLambda);
  }

  /** {@inheritDoc} */
  @Override
  public void revertMove() {
    orthogonalSpaceTempering.setLambda(currentLambda);
  }

  /**
   * Validate lambda is in the range [0 .. 1].
   *
   * <p>For discrete moves, set Lambda to the closest valid value [0, dL, 2dL, .. 1].
   *
   * @param lambda Input lambda value.
   * @return Validated lambda value.
   */
  public double validateLambda(double lambda) {
    lambda = max(0.0, min(lambda, 1.0));
    if (isContinuous) {
      return lambda;
    }
    double remainder = lambda % moveSize;
    if (remainder < moveSize / 2.0) {
      return max(0.0, lambda - remainder);
    } else {
      return min(lambda + (moveSize - remainder), 1.0);
    }
  }

  /**
   * Applies 0-1 mirroring conditions to lam + dL. Skips any moves where dL is greater than 1 or less
   * than -1, and skips 50% of moves from 0 or 1 (exact).
   *
   * @param lam Initial lambda.
   * @param dL Change in lambda.
   * @return Correctly mirrored lam + dL
   */
  private double mirror(double lam, double dL) {
    // Telescope to public static method because a public static method
    // may be useful in the future.
    return mirror(random, lam, dL);
  }

  /**
   * Pulls a delta-lambda from a continuous Gaussian distribution.
   *
   * @return A random Gaussian value with width of moveSize.
   */
  private double continuousMove() {
    // Draw a trial move from the distribution.
    return random.nextGaussian() * moveSize;
  }

  /**
   * Pulls a continuous lambda move of width moveSize.
   *
   * @return +/- moveSize (never 0)
   */
  private double discreteMove() {
    // Make a discrete move.
    double dL = moveSize;
    if (random.nextBoolean()) {
      dL = -moveSize;
    }
    return dL;
  }
}