package squidpony.squidmath; import squidpony.annotation.GwtIncompatible; import java.io.Serializable; import java.util.ArrayList; import java.util.Collection; import java.util.List; import java.util.Random; /** * A subclass of StatefulRNG (and thus RNG) that allows customizing many parts of the random number generation. * This is meant to be a more comprehensible version of the functionality present in RandomBias, and also for it to be * easier to use with methods that expect an RNG. * <br> * You can change the expected average for the values this produces, which uses the RandomBias.EXPONENTIAL distribution, * with all the caveats it has: it strongly favors either high or low values when the average gets especially high or * low, but it can essentially cover all averages between 0.0 and 1.0 (this class limits it to 0.1 and 0.9, so other * techniques can be used effectively). * <br> * You can also affect the "centrality" of random numbers, causing more to occur near the expected average (a bell curve * effect), or cause more near extreme ends of the random number spectrum. In practice, centrality changes are hard to * notice, but may be useful to simulate certain effects. An example of centrality changes in existing games include the * Nintendo title Advance Wars 2, where a brutish commander could increase the amount of damage his units dealt but also * suffered unpredictability; attacks could deal even more or much less damage than normal without any way to build * tactics around it. Square Enix's Final Fantasy XII also notably differentiated certain weapons (axes, hammers, and * "hand-cannons") from other similar options by making them deal less predictable damage. In both cases the connotation * is that more randomness is fitting for a brute-force approach to combat where pre-planned strategies are less * emphasized. It should also be noted that increasing the frequency of extreme results makes small bonuses to defense * or offense typically less useful, and small penalties less harmful. The opposite can be true for a carefully tuned * game where the most common results are tightly clustered, and most target numbers are just slightly above the * ordinary average. In tabletop games, 1d20 and 3d6 have the same average, but 1d20 is uniform, where 3d6 is clustered * around 10 and 11, each the result of 1/8 of rolls on their own and 1/4 together. This makes the case where a +1 bonus * to succeed changes the outcome on approximately 5% of 1d20 rolls, regardless of the required number to succeed if it * is less than 20. However, a +1 bonus matters on a variable portion of 3d6 rolls; if you become able to succeed on a * 10 or 11 where that was a failure before, the bonus applies approximately 12.5% of the time. Becoming able to succeed * on an 18 where that was a failure before is essentially worthless, affecting less than 0.5% of rolls. This property * of centralized results should be considered if game balance and/or the lethality of combat is important. One lengthy * stretch of extreme results by enemies that work against the favor of a player character generally result in a dead * player character, and RNGs that make extreme results more common may seem particularly cruel to players. * <br> * This generator sets a field, rawLatest, every time a random number is produced. This stores a pseudo-random double * between 0.0 (inclusive) and 1.0 (exclusive) that is not subject to the bias an expected average introduces, and is * close to uniformly distributed. You should expect rawLatest to be higher when higher numbers are returned from a * method like nextInt(), and lower when lower numbers are returned. This can be useful for rare effects that should not * be drastically more or less likely when slight changes are made to the expected average; if the expected average is * 0.65, many more random doubles from nextDouble() will be between 0.95 and 1.0 (probably more than 10% of random * numbers), but rawLatest will only be between 0.95 and 1.0 for close to 5% of all generations. * <br> * You can get and set the state this uses internally, and this is stored as a 64-bit long. * <br> * The choice of RandomnessSource doesn't really matter since this will always use a LightRNG internally. LightRNG is * the current best StatefulRandomness implementation, with excellent performance characteristics and few flaws, though * its relatively low period may sometimes be a detriment. * <br> * More customizations may be added in the future to the ones available currently. */ public class EditRNG extends StatefulRNG implements Serializable{ /** Used to tweak the generator toward high or low values. */ private double expected = 0.5; /** * When positive, makes the generator more likely to generate values close to the average (bell curve). * When zero (the default), makes no changes to the centering of values. * When negative, makes the generator swing more toward extremes rather than gravitate toward the average. * Values are typically between -100 and 100, but can have extreme weight and overshadow other parts of the RNG if * they go much higher than 200. */ private double centrality = 0.0; /** * The latest generated double, between 0.0 and 1.0, before changes for centrality and expected average. * Doubles are used to generate all random numbers this class produces, so be aware that calling getRandomElement() * will change this just as much as nextDouble(), nextInt(), or between() will. Primarily useful to obtain * uniformly-distributed random numbers that are related to the biased random numbers this returns as a main result, * such as to find when the last number generated was in the bottom 5% (less than 0.05, which could represent some * kind of critical failure or fumble) or top 10% (greater than or equal to 0.9, which could grant a critical * success or luck-based reward of some kind). */ public double rawLatest = 0.5; private static final long serialVersionUID = -2458726316853811777L; /** * Constructs an EditRNG with a pseudo-random seed from Math.random(). */ public EditRNG() { } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. */ public EditRNG(final long seed) { super(seed); } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. */ public EditRNG(final String seed) { super(seed); } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 */ public EditRNG(final long seed, double expected) { super(seed); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 */ public EditRNG(final String seed, double expected) { super(seed); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 * @param centrality if positive, makes results more likely to be near expected; if negative, the opposite. The * absolute value of centrality affects how centered results will be, with 0 having no effect */ public EditRNG(final long seed, double expected, double centrality) { super(seed); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); this.centrality = centrality; } /** * Construct a new EditRNG with the given seed. * * @param seed used to seed the default RandomnessSource. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 * @param centrality if positive, makes results more likely to be near expected; if negative, the opposite. The * absolute value of centrality affects how centered results will be, with 0 having no effect */ public EditRNG(final String seed, double expected, double centrality) { super(seed); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); this.centrality = centrality; } /** * Construct a new EditRNG with the given seed. * * @param rs the implementation used to generate random bits. */ public EditRNG(final RandomnessSource rs) { super(rs); } /** * Construct a new EditRNG with the given seed. * * @param rs the implementation used to generate random bits. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 */ public EditRNG(final RandomnessSource rs, double expected) { super(rs); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); } /** * Construct a new EditRNG with the given seed. * * @param rs the implementation used to generate random bits. * @param expected the expected average for random doubles, which will be capped between 0.1 and 0.9 * @param centrality if positive, makes results more likely to be near expected; if negative, the opposite. The * absolute value of centrality affects how centered results will be, with 0 having no effect */ public EditRNG(final RandomnessSource rs, double expected, double centrality) { super(rs); this.expected = Math.max(0.1, Math.min(0.89999994, expected)); this.centrality = centrality; } /** * Generate a random double, altered to try to match the expected average and centrality. * @return a double between 0.0 (inclusive) and 1.0 (exclusive) */ @Override public double nextDouble() { long l = random.nextLong(); double gen = (l & 0x1fffffffffffffL) * DOUBLE_UNIT, scatter = (l & 0xffffffL) * FLOAT_UNIT; rawLatest = 0.9999999999999999 - gen; gen = 0.9999999999999999 - Math.pow(gen, 1.0 / (1.0 - expected) - 1.0); if(centrality > 0) { scatter = 0.9999999999999999 - Math.pow(scatter, 1.0 / (1.0 - expected) - 1.0); gen = (gen * 100 + scatter * centrality) / (100 + centrality); } else if(centrality < 0) { scatter = Math.sin(scatter * Math.PI * 0.5); scatter *= scatter; if(expected >= 0.5) scatter = scatter * (1.0 - expected) * 2 + expected - (1.0 - expected); else scatter *= expected * 2; gen = (gen * 100 - scatter * centrality) / (100 - centrality); } return gen; } /** * This returns a random double between 0.0 (inclusive) and max (exclusive). * * @return a value between 0 (inclusive) and max (exclusive) */ @Override public double nextDouble(double max) { return nextDouble() * max; } /** * Returns a value from a even distribution from min (inclusive) to max * (exclusive). * * @param min the minimum bound on the return value (inclusive) * @param max the maximum bound on the return value (exclusive) * @return the found value */ @Override public double between(double min, double max) { return super.between(min, max); } /** * Returns a value between min (inclusive) and max (exclusive). * * The inclusive and exclusive behavior is to match the behavior of the * similar method that deals with floating point values. * * @param min the minimum bound on the return value (inclusive) * @param max the maximum bound on the return value (exclusive) * @return the found value */ @Override public int between(int min, int max) { return super.between(min, max); } @Override public long between(long min, long max) { return super.between(min, max); } /** * Returns the average of a number of randomly selected numbers from the * provided range, with min being inclusive and max being exclusive. It will * sample the number of times passed in as the third parameter. * * The inclusive and exclusive behavior is to match the behavior of the * similar method that deals with floating point values. * * This can be used to weight RNG calls to the average between min and max. * * @param min the minimum bound on the return value (inclusive) * @param max the maximum bound on the return value (exclusive) * @param samples the number of samples to take * @return the found value */ @Override public int betweenWeighted(int min, int max, int samples) { return super.betweenWeighted(min, max, samples); } /** * Returns a random element from the provided array and maintains object * type. * * @param <T> the type of the returned object * @param array the array to get an element from * @return the randomly selected element */ @Override public <T> T getRandomElement(T[] array) { return super.getRandomElement(array); } /** * Returns a random element from the provided list. If the list is empty * then null is returned. * * @param <T> the type of the returned object * @param list the list to get an element from * @return the randomly selected element */ @Override public <T> T getRandomElement(List<T> list) { return super.getRandomElement(list); } /** * Returns a random element from the provided ShortSet. If the set is empty * then an exception is thrown. * * <p> * Requires iterating through a random amount of the elements in set, so performance depends on the size of set but * is likely to be decent. This is mostly meant for internal use, the same as ShortSet. * </p> * @param set the ShortSet to get an element from * @return the randomly selected element */ public short getRandomElement(ShortSet set) { return super.getRandomElement(set); } /** * Returns a random element from the provided Collection, which should have predictable iteration order if you want * predictable behavior for identical RNG seeds, though it will get a random element just fine for any Collection * (just not predictably in all cases). If you give this a Set, it should be a LinkedHashSet or some form of sorted * Set like TreeSet if you want predictable results. Any List or Queue should be fine. Map does not implement * Collection, thank you very much Java library designers, so you can't actually pass a Map to this, though you can * pass the keys or values. If coll is empty, returns null. * * <p> * Requires iterating through a random amount of coll's elements, so performance depends on the size of coll but is * likely to be decent, as long as iteration isn't unusually slow. This replaces {@code getRandomElement(Queue)}, * since Queue implements Collection and the older Queue-using implementation was probably less efficient. * </p> * @param <T> the type of the returned object * @param coll the Collection to get an element from; remember, Map does not implement Collection * @return the randomly selected element */ public <T> T getRandomElement(Collection<T> coll) { return super.getRandomElement(coll); } /** * @return a value from the gaussian distribution */ @Override public synchronized double nextGaussian() { return super.nextGaussian(); } /** * Returns a random integer below the given bound, or 0 if the bound is 0 or * negative. * * @param bound the upper bound (exclusive) * @return the found number */ @Override public int nextInt(int bound) { if (bound <= 0) { return 0; } return (int)(nextDouble() * bound); } /** * Returns a random integer, which may be positive or negative. * @return A random int */ @Override public int nextInt() { return (int)((nextDouble() * 2.0 - 1.0) * 0x7FFFFFFF); } /** * Returns a random long, which may be positive or negative. * @return A random long */ @Override public long nextLong() { return (long)((nextDouble() * 2.0 - 1.0) * 0x7FFFFFFFFFFFFFFFL); } /** * Returns a random long below the given bound, or 0 if the bound is 0 or * negative. * * @param bound the upper bound (exclusive) * @return the found number */ @Override public long nextLong(long bound) { if (bound <= 0) { return 0; } return (long)(nextDouble() * bound); } /** * Gets the current expected average for this EditRNG. * @return the current expected average. */ public double getExpected() { return expected; } /** * Sets the expected average for random doubles this produces, which must always be between 0.1 and 0.9, and will be * set to 0.5 if an invalid value is passed. * @param expected the expected average to use, which should be 0.1 <= fairness < 0.9 */ public void setExpected(double expected) { if(expected < 0.0 || expected >= 1.0) this.expected = 0.5; else this.expected = expected; } /** * Gets the current centrality measure of this EditRNG. * Centrality has several possible effects: * When positive, makes the generator more likely to generate values close to the average (bell curve). * When zero (the default), makes no changes to the centering of values. * When negative, makes the generator swing more toward extremes rather than gravitate toward the average. * <br> * Values are typically between -100 and 100, but can have extreme weight and overshadow other parts of the RNG if * they go much higher than 200. * @return the current centrality */ public double getCentrality() { return centrality; } /** * Gets the current centrality measure of this EditRNG. * Centrality has several possible effects: * When positive, makes the generator more likely to generate values close to the average (bell curve). * When zero (the default), makes no changes to the centering of values. * When negative, makes the generator swing more toward extremes rather than gravitate toward the average. * <br> * Values are typically between -100 and 100, but can have extreme weight and overshadow other parts of the RNG if * they go much higher than 200. * @param centrality the new centrality measure to use */ public void setCentrality(double centrality) { this.centrality = centrality; } /** * * @param bits the number of bits to be returned * @return a random int of the number of bits specified. */ @Override public int next(int bits) { if(bits <= 0) return 0; if(bits > 32) bits = 32; return (int)(nextDouble() * (1L << bits)); } @Override public Random asRandom() { return super.asRandom(); } @Override @GwtIncompatible public <T> List<T> randomRotation(List<T> l) { return super.randomRotation(l); } @Override public <T> Iterable<T> getRandomStartIterable(List<T> list) { return super.getRandomStartIterable(list); } @Override public <T> T[] shuffle(T[] elements, T[] dest) { return super.shuffle(elements, dest); } @Override public <T> ArrayList<T> shuffle(Collection<T> elements) { return super.shuffle(elements); } @Override public float nextFloat() { return (float)nextDouble(); } @Override public boolean nextBoolean() { return nextDouble() >= 0.5; } @Override public RandomnessSource getRandomness() { return random; } @Override public void setRandomness(RandomnessSource random) { this.random = random; } /** * Gets a random portion of data (an array), assigns that portion to output (an array) so that it fills as much as * it can, and then returns output. Will only use a given position in the given data at most once; does this by * shuffling a copy of data and getting a section of it that matches the length of output. * * Based on http://stackoverflow.com/a/21460179 , credit to Vincent van der Weele; modifications were made to avoid * copying or creating a new generic array (a problem on GWT). * @param data an array of T; will not be modified. * @param output an array of T that will be overwritten; should always be instantiated with the portion length * @param <T> can be any non-primitive type. * @return an array of T that has length equal to output's length and may contain null elements if output is shorter * than data */ @Override public <T> T[] randomPortion(T[] data, T[] output) { return super.randomPortion(data, output); } /** * Gets a random portion of a List and returns it as a new List. Will only use a given position in the given * List at most once; does this by shuffling a copy of the List and getting a section of it. * * @param data a List of T; will not be modified. * @param count the non-negative number of elements to randomly take from data * @return a List of T that has length equal to the smaller of count or data.length */ @Override public <T> List<T> randomPortion(List<T> data, int count) { return super.randomPortion(data, count); } /** * Gets a random subrange of the non-negative ints from start (inclusive) to end (exclusive), using count elements. * May return an empty array if the parameters are invalid (end is less than/equal to start, or start is negative). * * @param start the start of the range of numbers to potentially use (inclusive) * @param end the end of the range of numbers to potentially use (exclusive) * @param count the total number of elements to use; will be less if the range is smaller than count * @return an int array that contains at most one of each number in the range */ @Override public int[] randomRange(int start, int end, int count) { return super.randomRange(start, end, count); } /** * Gets the latest "un-biased" random double used to produce the most recent (potentially) biased random number * generated for another method in this class, such as nextDouble(), between(), or getRandomElement(). This is a * double between 0.0 (inclusive) and 1.0 (exclusive). * @return the latest uniformly-distributed double before bias is added; between 0.0 and 1.0 (exclusive upper) */ public double getRawLatest() { return rawLatest; } /** * Creates a copy of this StatefulRNG; it will generate the same random numbers, given the same calls in order, as * this StatefulRNG at the point copy() is called. The copy will not share references with this StatefulRNG. * * @return a copy of this StatefulRNG */ @Override public RNG copy() { EditRNG next = new EditRNG(random.copy(), expected, centrality); next.rawLatest = rawLatest; return next; } /** * Get a long that can be used to reproduce the sequence of random numbers this object will generate starting now. * * @return a long that can be used as state. */ @Override public long getState() { return super.getState(); } /** * Sets the state of the random number generator to a given long, which will alter future random numbers this * produces based on the state. * * @param state a long, which typically should not be 0 (some implementations may tolerate a state of 0, however). */ @Override public void setState(long state) { super.setState(state); } @Override public String toString() { return "EditRNG{" + "expected=" + expected + ", centrality=" + centrality + ", Randomness Source=" + random + '}'; } @Override public boolean equals(Object o) { if (this == o) return true; if (o == null || getClass() != o.getClass()) return false; if (!super.equals(o)) return false; EditRNG editRNG = (EditRNG) o; if (Double.compare(editRNG.expected, expected) != 0) return false; return Double.compare(editRNG.centrality, centrality) == 0; } @Override public int hashCode() { int result = super.hashCode() * 31; long temp; temp = NumberTools.doubleToLongBits(expected); result += (int) (temp ^ (temp >>> 32)); temp = NumberTools.doubleToLongBits(centrality); result = 31 * result + (int) (temp ^ (temp >>> 32)); return result; } /** * Shuffle an array using the Fisher-Yates algorithm. Not GWT-compatible; use the overload that takes two arrays. * <br> * https://en.wikipedia.org/wiki/Fisher%E2%80%93Yates_shuffle * * @param elements an array of T; will not be modified * @return a shuffled copy of elements */ @GwtIncompatible @Override public <T> T[] shuffle(T[] elements) { return super.shuffle(elements); } /** * Generates a random permutation of the range from 0 (inclusive) to length (exclusive). * Useful for passing to OrderedMap or OrderedSet's reorder() methods. * * @param length the size of the ordering to produce * @return a random ordering containing all ints from 0 to length (exclusive) */ @Override public int[] randomOrdering(int length) { return super.randomOrdering(length); } /** * Returns a random non-negative integer below the given bound, or 0 if the bound is 0. * Uses a slightly optimized technique. This method is considered "hasty" since * it should be faster than nextInt() doesn't check for "less-valid" bounds values. It also * has undefined behavior if bound is negative, though it will probably produce a negative * number (just how negative is an open question). * * @param bound the upper bound (exclusive); behavior is undefined if bound is negative * @return the found number */ @Override public int nextIntHasty(int bound) { return (int)(nextDouble() * bound); } }