×

Simulating large Gaussian random vectors subject to inequality constraints by Gibbs sampling. (English) Zbl 1322.65030

Summary: The Gibbs sampler is an iterative algorithm used to simulate Gaussian random vectors subject to inequality constraints. This algorithm relies on the fact that the distribution of a vector component conditioned by the other components is Gaussian, the mean and variance of which are obtained by solving a kriging system. If the number of components is large, kriging is usually applied with a moving search neighborhood, but this practice can make the simulated vector not reproduce the target correlation matrix. To avoid these problems, variations of the Gibbs sampler are presented. The conditioning to inequality constraints on the vector components can be achieved by simulated annealing or by restricting the transition matrix of the iterative algorithm. Numerical experiments indicate that both approaches provide realizations that reproduce the correlation matrix of the Gaussian random vector, but some conditioning constraints may not be satisfied when using simulated annealing. On the contrary, the restriction of the transition matrix manages to satisfy all the constraints, although at the cost of a large number of iterations.

MSC:

65C60 Computational problems in statistics (MSC2010)
62M40 Random fields; image analysis

Software:

tmvtnorm
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Amit Y, Grenander U (1991) Comparing sweep strategies for stochastic relaxation. J Multivar Anal 37(2):197-222 · Zbl 0735.60035 · doi:10.1016/0047-259X(91)90080-L
[2] Armstrong M, Galli A, Beucher H, Le Loc’h G, Renard D, Doligez B, Eschard R, Geffroy F (2011) Plurigaussian simulations in geosciences. Springer, Berlin · doi:10.1007/978-3-642-19607-2
[3] Arroyo D, Emery X, Peláez M (2012) An enhanced Gibbs sampler algorithm for non-conditional simulation of Gaussian random vectors. Comput Geosci 46:138-148 · doi:10.1016/j.cageo.2012.04.011
[4] Bárdossy A (2006) Copula-based geostatistical models for groundwater quality parameters. Water Resour Res 42:W11416
[5] Chilès JP, Delfiner P (2012) Geostatistics: modeling spatial uncertainty. Wiley, New York · Zbl 1256.86007 · doi:10.1002/9781118136188
[6] Dubrule O (1983) Cross validation of kriging in a unique neighborhood. Math Geol 15(6):687-699 · doi:10.1007/BF01033232
[7] Emery X (2005) Conditional simulation of random fields with bivariate gamma isofactorial distributions. Math Geol 37(4):419-445 · Zbl 1122.86307 · doi:10.1007/s11004-005-5956-0
[8] Emery X (2007a) Using the Gibbs sampler for conditional simulation of Gaussian-based random fields. Comput Geosci 33(4):522-537 · doi:10.1016/j.cageo.2006.08.003
[9] Emery X (2007b) Simulation of geological domains using the plurigaussian model: new developments and computer programs. Comput Geosci 33(9):1189-1201 · doi:10.1016/j.cageo.2007.01.006
[10] Emery X (2008) Statistical tests for validating geostatistical simulation algorithms. Comput Geosci 34(11):1610-1620 · doi:10.1016/j.cageo.2007.12.012
[11] Freulon, X.; Armstrong, M. (ed.); Dowd, PA (ed.), Conditional simulation of a Gaussian random vector with nonlinear and/or noisy observations, 57-71 (1994), Dordrecht · doi:10.1007/978-94-015-8267-4_5
[12] Freulon, X.; Fouquet, C.; Soares, A. (ed.), Conditioning a Gaussian model with inequalities, 201-212 (1993), Dordrecht · doi:10.1007/978-94-011-1739-5_17
[13] Galli A, Gao H (2001) Rate of convergence of the Gibbs sampler in the Gaussian case. Math Geol 33(6):653-677 · Zbl 1011.86008 · doi:10.1023/A:1011094131273
[14] Geweke, JF; Keramidas, EM (ed.), Efficient simulation from the multivariate normal and student-t distributions subject to linear constraints and the evaluation of constraint probabilities, 571-578 (1991), Fairfax
[15] Journel AG (1986) Constrained interpolation and qualitative information—the soft kriging approach. Math Geol 18(3):269-286 · doi:10.1007/BF00898032
[16] Kotecha, JH; Djuric, PM, Gibbs sampling approach for generation of truncated multivariate Gaussian random variables, 1757-1760 (1999), Washington
[17] Lantuéjoul C (2002) Geostatistical simulation: models and algorithms. Springer, Berlin · Zbl 0990.86007 · doi:10.1007/978-3-662-04808-5
[18] Lantuéjoul, C.; Desassis, N., Simulation of a Gaussian random vector: a propagative version of the Gibbs sampler, Oslo
[19] Loc’h, G.; Beucher, H.; Galli, A.; Doligez, B.; Heresim Group; Omre, H. (ed.); Thomassen, P. (ed.), Improvement in the truncated Gaussian method: combining several Gaussian functions (1994), Røros
[20] Levine RA, Casella G (2006) Optimizing random scan Gibbs samplers. J Multivar Anal 97:2071-2100 · Zbl 1101.62087 · doi:10.1016/j.jmva.2006.05.008
[21] Matheron, G.; Beucher, H.; Fouquet, C.; Galli, A.; Guérillot, D.; Ravenne, C., Conditional simulation of the geometry of fluvio-deltaic reservoirs, Dallas
[22] Roberts, GO; Gilks, WR (ed.); Richardson, S. (ed.); Spiegelhalter, DJ (ed.), Markov chain concepts related to sampling algorithms, 45-57 (1996), Boca Raton · Zbl 0839.62078 · doi:10.1007/978-1-4899-4485-6_3
[23] Roberts GO, Sahu SK (1997) Updating schemes, correlation structure, blocking and parameterization for the Gibbs sampler. J R Stat Soc B 59(2):291-317 · Zbl 0886.62083 · doi:10.1111/1467-9868.00070
[24] Tierney L (1994) Markov chains for exploring posterior distributions. Ann Stat 22(4):1701-1762 · Zbl 0829.62080 · doi:10.1214/aos/1176325750
[25] Wilhelm S, Manjunath BG (2011) Tmvtnorm: truncated multivariate normal and Student t distribution. R package version 1.4-8. http://cran.r-project.org/package=tmvtnorm. Accessed 2 July 2013 · Zbl 0829.62080
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.