Hi, I have worked on the numerical stability problems (see MATH-718, MATH-785) of our continued fraction implementation, which is used in several other classes (Beta, Gamma).
As I was not really convinced with the scaling approach to overcome overflow and underflow situations which occur quite often, I tried out the modified Lentz-Thompson algorithm (as outlined in Numerical Recipes, but I found also other web references). This seems to solve all the described problems, but the question remains if this can be included without problems into CM. The algorithm description in NR is not in source code but on an higher level. Thomas --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org For additional commands, e-mail: dev-h...@commons.apache.org
