commons-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Phil Steitz (JIRA)" <j...@apache.org>
Subject [jira] [Updated] (MATH-877) Allow catagory data to be provided as SummaryStatistics in OneWayAnova
Date Tue, 09 Oct 2012 15:20:02 GMT

     [ https://issues.apache.org/jira/browse/MATH-877?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Phil Steitz updated MATH-877:
-----------------------------

    Attachment: OneWayAnova.java

Attached by Peter Andrews in 9 Oct 2012 post to dev@commons.apache.org
                
> Allow catagory data to be provided as SummaryStatistics in OneWayAnova
> ----------------------------------------------------------------------
>
>                 Key: MATH-877
>                 URL: https://issues.apache.org/jira/browse/MATH-877
>             Project: Commons Math
>          Issue Type: Improvement
>    Affects Versions: 3.0
>            Reporter: Phil Steitz
>             Fix For: 3.2
>
>         Attachments: OneWayAnova.java
>
>
> From a post on 9 Oct, 2012 to dev@commons.apache.org from Peter Andrews
> {quote}
> I have implemented a version of OneWayAnova which uses SummaryStatistics rather than
requiring the array of doubles.
> I know that there is a whole process for submitting code but I am under a deadline so
sending this email with the code is all I am going to do.
> A couple of notes:
>     I would have maintained the same signature structure simply adding methods that supported
Collection<SummaryStatistics> but that creates an error because it has the same Signature
due to type erasure on Collection<double[]>.
>     I did not see why certain cases such as having only one array or number of elements
< 2 should necessarily throw exceptions -- the math still works out and gives reasonable
results so given point #1, I differentiated the function signatures by adding a boolean to
optionally skip the checks and therefore not not throw DimensionMismatchException.
> Minor note: I understand that you cannot change it at this point but there is no reason
for this class not to abstract with static methods since it has no state.
> Do with this what you will.
> Thanks for the wonderful tools you provide.
> Peter Andrews
> {quote}

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira

Mime
View raw message