|
ABSTRACT
This paper describes a method of generating gamma variates that appears to be less costly than Wallace's recently suggested method. For large shape parameter &agr;; the cost of computation is proportional to √&agr;, whereas Wallace's method is proportional to &agr;. Experimentation by Robinson and Lewis indicates that for small &agr; the method suggested here also dominates methods recently suggested by Dieter and Ahrens, albeit those methods dominate for large &agr;. The method suggested here uses the rejection technique.
REFERENCES
Note: OCR errors may be found in this Reference List extracted from the full text article. ACM has opted to expose the complete List rather than only correct and linked references.
| |
1
|
Dieter, V. and Ahrens, J.H. Acceptance-rejection techniques for sampling from the gamma and beta distributions. Tech. Rep. No. 83, Dep. of Statistics, Stanford U., 1974.
|
| |
2
|
Robinson, D.W., and Lewis, P.A.W. Generating gamma and Cauchy random variables: an extension to the Naval Postgraduate School random number package. Naval Postgraduate School, Monterey, Calif., 1975.
|
 |
3
|
|
|