DO NOT REPLY TO THIS EMAIL, BUT PLEASE POST YOUR BUG RELATED COMMENTS THROUGH THE WEB INTERFACE AVAILABLE AT <http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20279>. ANY REPLY MADE TO THIS MESSAGE WILL NOT BE COLLECTED AND INSERTED IN THE BUG DATABASE.
http://nagoya.apache.org/bugzilla/show_bug.cgi?id=20279 [math] chi-squared distribution (part of the distribution collection) ------- Additional Comments From [EMAIL PROTECTED] 2003-05-28 01:12 ------- I like the generic CDF inversion capability provided by AbstractContinousDistribution, and the ContinuousDistribution interface. Personally, I do not see the implementation as over-engineered for what it does and if we want to tackle the critical value problem for statistical tests generically, I can't see a significantly simpler or more flexible way to do it. It would be nice to extend/modify/replace TestStatistic with a hypothesis- testing capability using the distribution framework to compute critical values for t (need to add this distribution), chi-square and maybe F. We will also obviously have to fill out the docs on the framework. What will likely be most useful to applications are the distribution implementations and any inferential machinery that we build on top of them. I guess there is no postponing the packaging discussion. stat is sort of obvious and I guess stat.distribution makes sense, assuming we want to move forward with the framework (which I personally think is a good idea). In an earlier post, I had designated something like "analysis" or "numerical", which would be a good home for the rootfinding stuff. Probably should take this to the list. --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
