[
https://issues.apache.org/jira/browse/STATISTICS-14?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16854120#comment-16854120
]
Alex D Herbert commented on STATISTICS-14:
------------------------------------------
Q. Why the {{bigdecimal}} package name?
Is the intension to have {{org.apache.commons.statistics.descriptive}} only
support {{double}}? These statistics apply to any number but at least matching
the JDK SummaryStatistics classes they could support {{double}}, {{long}} and
{{int}} to be useable in the specialised stream implementations for those types.
I don't see why this class cannot be in
{{org.apache.commons.statistics.descriptive}}.
Note that there could also be a specialisation for {{BigInteger}}. This would
warrant {{org.apache.commons.statistics.biginteger.descriptive}} under this
package naming convention.
Would it not be more consistent with the JDK primitive specialisations to have
all descriptive statistics prefixed with the primitive/BigNumber name and be in
the same package? E.g.
{code:java}
// For a single descriptive statistic
IntMean
LongMean
DoubleMean
BigDecimalMean
BigIntegerMean
// For functionality matching the JDK statistics implementations
BigDecimalStatistics
BigIntegerStatistics
{code}
A negative is that it would make the package quite crowded.
> BigDecimalStatistics
> --------------------
>
> Key: STATISTICS-14
> URL: https://issues.apache.org/jira/browse/STATISTICS-14
> Project: Apache Commons Statistics
> Issue Type: New Feature
> Reporter: Aleksander Ściborek
> Priority: Major
>
> The idea is to create a util class for reducing stream of BigDecimal objects
> and then getting an object which contains statistics such as
> 1)Min and max value
> 2)Average
> 3)Total sum
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
