Spectral test

Abstract: Combining parallel multiple recursive sequences provides an efficient way of implementing random number generators with long periods and good structural properties. Such generators are statistically more robust than simple linear congruential generators that fit into a computer word. We made extensive computer searches for good parameter sets, with respect to the spectral test, for combined multiple recursive generators of different sizes. We also compare different implementations and give a specific code in C that is faster than previous implementations of similar generators.

Abstract: From Probabilistic Diophantine Approximation to Quadratic Fields.- 1 Part I: Super Irregularity.- 2 Part II: Probabilistic Diophantine Approximation.- 2.1 Local Case: Inhomogeneous Pell Inequalities - Hyperbolas.- 2.2 Beyond Quadratic Irrationals.- 2.3 Global Case: Lattice Points in Tilted Rectangles.- 2.4 Simultaneous Case.- 3 Part III: Quadratic Fields and Continued Fractions.- 3.1 Cesaro Mean of % MathType!MTEF!2!1!+- % feaagCart1ev2aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn % hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqtubsr % 4rNCHbGeaGqiVu0Je9sqqrpepC0xbbL8F4rqqrFfpeea0xe9Lq-Jc9 % vqaqpepm0xbba9pwe9Q8fs0-yqaqpepae9pg0FirpepeKkFr0xfr-x % fr-xb9adbaqaaeGaciGaaiaabeqaamaabaabaaGcbaaeaaaaaaaaa8 % qadaaeabqaamaacmaabaGaamOBaiabeg7aH9aadaahaaWcbeqaa8qa % caaIXaGaai4laiaaikdaaaaakiaawUhacaGL9baaaSqabeqaniabgg % HiLdaaaa!3F6B!$$ \sum {\left\{ {n{\alpha ^{1/2}}} \right\}} $$ and Quadratic Fields.- 3.2 Hardy-Littlewood Lemma 14.- 4 Part IV: Class Number One Problems.- 4.1 An Attempt to Reduce the Yokoi's Conjecture to a Finite Amount of Computation.- 5 Part V: Cesaro Mean of % MathType!MTEF!2!1!+- % feaagCart1ev2aaatCvAUfeBSjuyZL2yd9gzLbvyNv2CaerbuLwBLn % hiov2DGi1BTfMBaeXatLxBI9gBaerbd9wDYLwzYbItLDharqqtubsr % 4rNCHbGeaGqiVu0Je9sqqrpepC0xbbL8F4rqqrFfpeea0xe9Lq-Jc9 % vqaqpepm0xbba9pwe9Q8fs0-yqaqpepae9pg0FirpepeKkFr0xfr-x % fr-xb9adbaqaaeGaciGaaiaabeqaamaabaabaaGcqaaaaaaaaaWdbe % aadaaeqaWdaeaapeWaaeWaa8aabaWdbmaacmaapaqaa8qacaWGUbGa % eqySdegacaGL7bGaayzFaaGaeyOeI0IaaGymaiaac+cacaaIYaaaca % GLOaGaayzkaaaal8aabaWdbiaad6gaaeqaniabggHiLdaaaa!42C9!$$ \sum olimits_n {\left( {\left\{ {n\alpha } \right\} - 1/2} \right)} $$.- 6 References.- On the Assessment of Random and Quasi-Random Point Sets.- 1 Introduction.- 2 Chapter for the Practitioner.- 2.1 Assessing RNGs.- 2.2 Correlation Analysis for RNGs I.- 2.3 Correlation Analysis for RNGs II.- 2.4 Theory vs. Practice I: Leap-Frog Streams.- 2.5 Theory vs. Practice II: Parallel Monte Carlo Integration.- 2.6 Assessing LDPs.- 2.7 Good Lattice Points.- 2.8 GLPs vs. (tms)-Nets.- 2.9 Conclusion.- 3 Mathematical Preliminaries.- 3.1 Haar and Walsh Series.- 3.2 Integration Lattices.- 4 Uniform Distribution Modulo One.- 4.1 The Definition of Uniformly Distributed Sequences.- 4.2 Weyl Sums and Weyl's Criterion.- 4.3 Remarks.- 5 The Spectral Test.- 5.1 Definition.- 5.2 Properties.- 5.3 Examples.- 5.4 Geometric Interpretation.- 5.5 Remarks.- 6 The Weighted Spectral Test.- 6.1 Definition.- 6.2 Examples and Properties.- 6.3 Remarks.- 7 Discrepancy.- 7.1 Definition.- 7.2 The Inequality of Erdos-Turan-Koksma.- 7.3 Remarks.- 8 Summary.- 9 Acknowledgements.- 10 References.- Lattice Rules: How Well Do They Measure Up?.- 1 Introduction.- 2 Some Basic Properties of Lattice Rules.- 3 A General Approach to Worst-Case and Average-Case Error Analysis.- 3.1 Worst-Case Quadrature Error for Reproducing Kernel Hilbert Spaces.- 3.2 A More General Worst-Case Quadrature Error Analysis.- 3.3 Average-Case Quadrature Error Analysis.- 4 Examples of Other Discrepancies.- 4.1 The ANOVA Decomposition.- 4.2 A Generalization ofP?(L) with Weights.- 4.3 The Periodic Bernoulli Discrepancy - Another Generalization ofP?(L).- 4.4 The Non-Periodic Bernoulli Discrepancy.- 4.5 The Star Discrepancy.- 4.6 The Unanchored Discrepancy.- 4.7 The Wrap-Around Discrepancy.- 4.8 The Symmetric Discrepancy.- 5 Shift-Invariant Kernels and Discrepancies.- 6 Discrepancy Bounds.- 6.1 Upper Bounds forP?(L).- 6.2 A Lower Bound onDF,?,1(P).- 6.3 Quadrature Rules with Different Weights.- 6.4 Copy Rules.- 7 Discrepancies of Integration Lattices and Nets.- 7.1 The Expected Discrepancy of Randomized (0ms)-Nets.- 7 2 Infinite Sequences of Embedded Lattices.- 8 Tractability of High Dimensional Quadrature.- 8.1 Quadrature in Arbitrarily High Dimensions.- 8.2 The Effective Dimension of an Integrand.- 9 Discussion and Conclusion.- 10 References.- Digital Point Sets: Analysis and Application.- 1 Introduction.- 2 The Concept and Basic Properties of Digital Point Sets.- 3 Discrepancy Bounds for Digital Point Sets.- 4 Special Classes of Digital Point Sets and Quality Bounds.- 5 Digital Sequences Based on Formal Laurent Series and Non-Archimedean Diophantine Approximation.- 6 Analysis of Pseudo-Random-Number Generators by Digital Nets.- 7 The Digital Lattice Rule.- 8 Outlook and Open Research Topics.- 9 References.- Random Number Generators: Selection Criteria and Testing.- 1 Introduction.- 2 Design Principles and Figures of Merit.- 2.1 A Roulette Wheel.- 2.2 Sampling from ?t.- 2.3 The Lattice Structure of MRG's.- 2.4 Equidistribution for Regular Partitions in Cubic Boxes.- 2.5 Other Measures of Divergence.- 3 Empirical Statistical Tests.- 3.1 What are the Good Tests?.- 3.2 Two-Level Tests.- 3.3 Collections of Empirical Tests.- 4 Examples of Empirical Tests.- 4.1 Serial Tests of Equidistribution.- 4.2 Tests Based on Close Points in Space.- 5 Collections of Small RNGs.- 5.1 Small Linear Congruential Generators.- 5.2 Explicit Inversive Congruential Generators.- 5.3 Compound Cubic Congruential Generators.- 6 Systematic Testing for Small RNGs.- 6.1 Serial Tests of Equidistribution for LCGs.- 6.2 Serial Tests of Equidistribution for Nonlinear Generators.- 6.3 A Summary of the Serial Tests Results.- 6.4 Close-Pairs Tests for LCGs.- 6.5 Close-Pairs Tests for Nonlinear Generators.- 6.6 A Summary of the Close-Pairs Tests Results.- 7 How Do Real-Life Generators Fare in These Tests?.- 8 Acknowledgements.- 9 References.- Nets, (ts)-Sequences, and Algebraic Geometry.- 1 Introduction.- 2 Basic Concepts.- 3 The Digital Method.- 4 Background on Algebraic Curves over Finite Fields.- 5 Construction of (ts)-Sequences.- 6 New Constructions of (tms)-Nets.- 7 New Algebraic Curves with Many Rational Points.- 8 References.- Financial Applications of Monte Carlo and Quasi-Monte Carlo Methods.- 1 Introduction.- 2 Monte Carlo Methods for Finance Applications.- 2.1 Preliminaries for Derivative Pricing.- 2.2 Variance Reduction Techniques.- 2.3 Caveats for Computer Implementation.- 3 Speeding Up by Quasi-Monte Carlo Methods.- 3.1 What are Quasi-Monte Carlo Methods?.- 3.2 Generalized Faure Sequences.- 3.3 Numerical Experiments.- 3.4 Discussions.- 4 Future Topics.- 4.1 Monte Carlo Simulations for American Options.- 4.2 Research Issues Related to Quasi-Monte Carlo Methods.- 5 References.

Abstract: In this paper we review some statistical tests included in the NIST SP 800-22 suite, which is a collection of tests for the evaluation of both true-random (physical) and pseudorandom (algorithmic) number generators for cryptographic applications. The output of these tests is the so-called p-value which is a random variable whose distribution converges to the uniform distribution in the interval [0,1] when testing an increasing number of samples from an ideal generator. Here, we compute the exact non-asymptotic distribution of p-values produced by few of the tests in the suite, and propose some computation-friendly approximations. This allows us to explain why intensive testing produces false-positives with a probability much higher than the expected one when considering asymptotic distribution instead of the true one. We also propose a new approximation for the Spectral Test reference distribution, which is more coherent with experimental results.