Series of generated values pass the TestU01 BigCrush and PractRand test suites + * that measure independence and uniformity properties of random number generators. + * + *
The class {@code Xoroshiro128StarStar} uses the {@code xoroshiro128} algorithm, + * version 1.0 (parameters 24, 16, 37), with the "**" scrambler (a mixing function). + * Its state consists of two {@code long} fields {@code x0} and {@code x1}, + * which can take on any values provided that they are not both zero. + * The period of this generator is 2128-1. + * + *
The 64-bit values produced by the {@code nextLong()} method are equidistributed. + * To be precise, over the course of the cycle of length 2128-1, + * each nonzero {@code long} value is generated 264 times, + * but the value 0 is generated only 264-1 times. + * The values produced by the {@code nextInt()}, {@code nextFloat()}, and {@code nextDouble()} + * methods are likewise equidistributed. + * + *
In fact, the 64-bit values produced by the {@code nextLong()} method are 2-equidistributed. + * To be precise: consider the (overlapping) length-2 subsequences of the cycle of 64-bit + * values produced by {@code nextLong()} (assuming no other methods are called that would + * affect the state). There are 2128-1 such subsequences, and each subsequence, + * which consists of 2 64-bit values, can have one of 2128 values. Of those + * 2128 subsequence values, each one is generated exactly once over the course + * of the entire cycle, except that the subsequence (0, 0) never appears. + * The values produced by the {@code nextInt()}, {@code nextFloat()}, and {@code nextDouble()} + * methods are likewise 2-equidistributed, but note that that the subsequence (0, 0) + * can also appear (but occurring somewhat less frequently than all other subsequences), + * because the values produced by those methods have fewer than 64 randomly chosen bits. + * + *
Instances {@code Xoroshiro128StarStar} are not thread-safe. + * They are designed to be used so that each thread as its own instance. + * The methods {@link #jump} and {@link #leap} and {@link #jumps} and {@link #leaps} + * can be used to construct new instances of {@code Xoroshiro128StarStar} that traverse + * other parts of the state cycle. + * + *
Instances of {@code Xoroshiro128StarStar} are not cryptographically
+ * secure. Consider instead using {@link java.security.SecureRandom}
+ * in security-sensitive applications. Additionally,
+ * default-constructed instances do not use a cryptographically random
+ * seed unless the {@linkplain System#getProperty system property}
+ * {@code java.util.secureRandomSeed} is set to {@code true}.
+ *
+ * @author Guy Steele
+ * @author Doug Lea
+ * @since 1.8
+ */
+public final class Xoroshiro128StarStar implements LeapableRng {
+
+ /*
+ * Implementation Overview.
+ *
+ * This is an implementation of the xoroshiro128** algorithm written
+ * in 2016 by David Blackman and Sebastiano Vigna (vigna@acm.org),
+ * and updated with improved parameters in 2018.
+ * See http://xoshiro.di.unimi.it and these two papers:
+ *
+ * Sebastiano Vigna. 2016. An Experimental Exploration of Marsaglia's
+ * xorshift Generators, Scrambled. ACM Transactions on Mathematical
+ * Software 42, 4, Article 30 (June 2016), 23 pages.
+ * https://doi.org/10.1145/2845077
+ *
+ * David Blackman and Sebastiano Vigna. 2018. Scrambled Linear
+ * Pseudorandom Number Generators. Computing Research Repository (CoRR).
+ * http://arxiv.org/abs/1805.01407
+ *
+ * The jump operation moves the current generator forward by 2*64
+ * steps; this has the same effect as calling nextLong() 2**64
+ * times, but is much faster. Similarly, the leap operation moves
+ * the current generator forward by 2*96 steps; this has the same
+ * effect as calling nextLong() 2**96 times, but is much faster.
+ * The copy method may be used to make a copy of the current
+ * generator. Thus one may repeatedly and cumulatively copy and
+ * jump to produce a sequence of generators whose states are well
+ * spaced apart along the overall state cycle (indeed, the jumps()
+ * and leaps() methods each produce a stream of such generators).
+ * The generators can then be parceled out to other threads.
+ *
+ * File organization: First the non-public methods that constitute the
+ * main algorithm, then the public methods. Note that many methods are
+ * defined by classes {@code AbstractJumpableRng} and {@code AbstractRng}.
+ */
+
+ /* ---------------- static fields ---------------- */
+
+ /**
+ * The seed generator for default constructors.
+ */
+ private static final AtomicLong defaultGen = new AtomicLong(RngSupport.initialSeed());
+
+ /*
+ * The period of this generator, which is 2**128 - 1.
+ */
+ private static final BigInteger thePeriod =
+ BigInteger.ONE.shiftLeft(128).subtract(BigInteger.ONE);
+
+ /* ---------------- instance fields ---------------- */
+
+ /**
+ * The per-instance state.
+ * At least one of the two fields x0 and x1 must be nonzero.
+ */
+ private long x0, x1;
+
+ /* ---------------- constructors ---------------- */
+
+ /**
+ * Basic constructor that initializes all fields from parameters.
+ * It then adjusts the field values if necessary to ensure that
+ * all constraints on the values of fields are met.
+ *
+ * @param x0 first word of the initial state
+ * @param x1 second word of the initial state
+ */
+ public Xoroshiro128StarStar(long x0, long x1) {
+ this.x0 = x0;
+ this.x1 = x1;
+ // If x0 and x1 are both zero, we must choose nonzero values.
+ if ((x0 | x1) == 0) {
+ // At least one of the two values generated here will be nonzero.
+ this.x0 = RngSupport.mixStafford13(x0 += RngSupport.GOLDEN_RATIO_64);
+ this.x1 = (x0 += RngSupport.GOLDEN_RATIO_64);
+ }
+ }
+
+ /**
+ * Creates a new instance of {@code Xoroshiro128StarStar} using the
+ * specified {@code long} value as the initial seed. Instances of
+ * {@code Xoroshiro128StarStar} created with the same seed in the same
+ * program generate identical sequences of values.
+ *
+ * @param seed the initial seed
+ */
+ public Xoroshiro128StarStar(long seed) {
+ // Using a value with irregularly spaced 1-bits to xor the seed
+ // argument tends to improve "pedestrian" seeds such as 0 or
+ // other small integers. We may as well use SILVER_RATIO_64.
+ //
+ // The x values are then filled in as if by a SplitMix PRNG with
+ // GOLDEN_RATIO_64 as the gamma value and Stafford13 as the mixer.
+ this(RngSupport.mixStafford13(seed ^= RngSupport.SILVER_RATIO_64),
+ RngSupport.mixStafford13(seed + RngSupport.GOLDEN_RATIO_64));
+ }
+
+ /**
+ * Creates a new instance of {@code Xoroshiro128StarStar} that is likely to
+ * generate sequences of values that are statistically independent
+ * of those of any other instances in the current program execution,
+ * but may, and typically does, vary across program invocations.
+ */
+ public Xoroshiro128StarStar() {
+ // Using GOLDEN_RATIO_64 here gives us a good Weyl sequence of values.
+ this(defaultGen.getAndAdd(RngSupport.GOLDEN_RATIO_64));
+ }
+
+ /**
+ * Creates a new instance of {@code Xoroshiro128StarStar} using the specified array of
+ * initial seed bytes. Instances of {@code Xoroshiro128StarStar} created with the same
+ * seed array in the same program execution generate identical sequences of values.
+ *
+ * @param seed the initial seed
+ */
+ public Xoroshiro128StarStar(byte[] seed) {
+ // Convert the seed to 2 long values, which are not both zero.
+ long[] data = RngSupport.convertSeedBytesToLongs(seed, 2, 2);
+ long x0 = data[0], x1 = data[1];
+ this.x0 = x0;
+ this.x1 = x1;
+ }
+
+ /* ---------------- public methods ---------------- */
+
+ public Xoroshiro128StarStar copy() { return new Xoroshiro128StarStar(x0, x1); }
+
+/*
+
+To the extent possible under law, the author has dedicated all copyright
+and related and neighboring rights to this software to the public domain
+worldwide. This software is distributed without any warranty.
+
+See