Fast simulation of truncated Gaussian distributions
- 906 Downloads
We consider the problem of simulating a Gaussian vector X, conditional on the fact that each component of X belongs to a finite interval [a i ,b i ], or a semi-finite interval [a i ,+∞). In the one-dimensional case, we design a table-based algorithm that is computationally faster than alternative algorithms. In the two-dimensional case, we design an accept-reject algorithm. According to our calculations and numerical studies, the acceptance rate of this algorithm is bounded from below by 0.5 for semi-finite truncation intervals, and by 0.47 for finite intervals. Extension to three or more dimensions is discussed.
Unable to display preview. Download preview PDF.
- Abramowitz, M., Stegun, I.: Handbook of Mathematical Functions with Formulas, Graphs, and Mathematical Table. Dover, New York (1965) Google Scholar
- Ahrens, J.: Sampling from general distributions by suboptimal division of domains. Grazer Math. Ber. 20(319) (1993) Google Scholar
- Blair, J., Edwards, C., Johnson, J.: Rational Chebyshev approximations for the inverse of the error function. Math. Comput. (1976) Google Scholar
- Geweke, J.: Efficient simulation from the multivariate normal and Student-t distributions subject to linear constraints and the evaluation of constraint probabilities. Comput. Sci. Stat. 23, 571–578 (1991) Google Scholar
- Hörmann, W., Leydold, J.: A note on perfect slice sampling. Technical Report 29, Dept. Stats. Maths. Wirtschaftsuniv (2006) Google Scholar
- Marsaglia, G., Tsang, W.: The Ziggurat method for generating random variables. J. Stat. Soft. 5(8) (2000) Google Scholar
- Wichura, M.: Algorithm AS 241: The percentage points of the normal distribution. Appl. Stat., 477–484 (1988) Google Scholar
- Zaman, A.: Generating random numbers from a unimodal density by cutting corners. Unpublished manuscript (1996) Google Scholar