[
https://issues.apache.org/jira/browse/MATH-1038?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13877650#comment-13877650
]
Phil Steitz commented on MATH-1038:
-----------------------------------
I think "interval" is fine as subpackage and we may as well move the
ConfidenceInterval and abstract base class there as well. The question is
where to place it. One alternative to consider is to make interval a peer of
inference. That would set us up to eliminate "inference" in 4.0, effectively
splitting it into "interval" and "test". It would be logical to keep
"inference" as the parent for "test" and "interval" (and maybe somewhere to put
Anova), but my HO is flatter is better when it comes to finding stuff, so I
would favor stat.interval, stat.test and ditch .inference in 4.0. That means
now we create stat.interval and put all of the confidence interval stuff there.
> Add methods to find confidence intervals for the mean of a binomial
> distribution
> --------------------------------------------------------------------------------
>
> Key: MATH-1038
> URL: https://issues.apache.org/jira/browse/MATH-1038
> Project: Commons Math
> Issue Type: Wish
> Affects Versions: 3.2
> Reporter: Thorsten Schäfer
> Assignee: Phil Steitz
> Priority: Minor
> Fix For: 3.3
>
> Attachments: binomialConfidenceInterval.patch,
> binomialConfidenceInterval.patch,
> binomial_confidence_interval_refactored.patch
>
>
--
This message was sent by Atlassian JIRA
(v6.1.5#6160)
