[
https://issues.apache.org/jira/browse/MATH-1185?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14265294#comment-14265294
]
Gilles commented on MATH-1185:
------------------------------
bq. The original wish was for greater precision for cumProbability but this
assumes there are reliable algorithms that achieve this
It's not a problem of the algorithm being more or less accurate; it's one of
limited precision.
As in the example mentioned by Luc ("expm1"), "oneMinusCumulativeProbability"
could leverage a more accurate algorithm (for large input), but
"cumulativeProbability" can't.
It seems that the library you refer to below indeed computes q = 1 - p, IIUC.
Unfortunately it looks like it would be nightmare to port: there are almost as
many "goto"s as there are lines of code :(
> Tail probability drops to zero beyond 10e-17 ?
> ----------------------------------------------
>
> Key: MATH-1185
> URL: https://issues.apache.org/jira/browse/MATH-1185
> Project: Commons Math
> Issue Type: Wish
> Affects Versions: 3.3, 3.4
> Reporter: Sriram Natarajan
> Priority: Minor
>
> This could be a simple question, In which case I can expect a clarification.
> If this is the wrong place to post such a question, let me know.
> OS: Windows 8.1, Java 1.8.0_25
> ChiSquaredDistribution chisq = new ChiSquaredDistribution(23)
> 1.0 - chisq.cumulativeProbability(130) is 1.1102230246251565E-16
> 1.0 - chisq.cumulativeProbability(131) is 0.0
> Am pretty sure it is not a formatting issue. Is there a reason why the tail
> drops to zero at this point?
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
