|
ABSTRACT
Algorithm 708 (BRATIO) was run on 2730 test cases. Comparison of these results with the results from an algorithm using a continued fraction of Tretter and Walster were performed using a high-precision version of the latter algorithm implemented in Maple. Accuracy of BRATIO ranged from 9.64 significant digits to a full machine double-precision, 15.65 significant digits, with the lower value occurring when a was nearly equal to b, and a was large.
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
|
|
 |
2
|
|
 |
3
|
|
| |
4
|
KALBFLEISH, J. D., AND PRENTICE, R. L. 1980. The Statistical Analysts of Failure Ttme Data. John Wiley and Sons Inc., New York.
|
| |
5
|
STATISTICAL SCIENCES, INC. 1991. S~Plus Version 3.0. Statistical Sciences, Inc., Seattle, Wash. (Computer Program).
|
| |
6
|
TRETTER, M. J., AND WALSTER, G. W. 1980. Analytic subtraction applied to the incomplete gamma and beta functions. SIAM J. Sci. Stat. Comput. 1,321 326.
|
| |
7
|
TRETTER, M. J., AND WALSTER, G. W. 1979. Continued fractions for the incomplete beta function: Additions and corrections. Ann. Star. 7, 2, 462 465.
|
| |
8
|
WATERLOO MAPLE SOFTWARE. 1990. Maple V. Waterloo Maple Software, Waterloo, Ont. (Computer Program).
|
CITED BY
|
|
Barry W. Brown , Lawrence B. Levy , James Lovato , Kathy Russell , Floyd M. Spears, Algorithm 762; LLDRLF, log-likelihood and some derivatives for log-F models, ACM Transactions on Mathematical Software (TOMS), v.22 n.3, p.372-382, Sept. 1996
|
|