[math] checkstyle.wrap.operator = ignore

["Mark R. Diggory" <md...@latte.harvard.edu>]

I'm concerned that this property does not actually do what we think it 
is doing. "ignore" is used because there were problems using the "eol" 
string in the flat property files. Anyways this forces the checkstyle 
parser to require all operators to be at the end of the lines vs 
beginning of lines. There is no way to ignore if its at the end or 
beginning of the line.

Most autoformaters (Eclipse in my case) format according the checkstyle 
spec such that all operators fall at the beginnings (I don't see a way 
to change this in Eclipse at least). I suspect it covers more IDE cases 
to have this set property removed so it defaults to the beginning of the 
line.

-Mark


---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org

[math] Thoughts on efficiency in ListUnivariates

["Mark R. Diggory" <md...@latte.harvard.edu>]

The new ListUnivariateImpl gives an opportunity to ponder a couple 
different approaches and their efficiency issues.

Current approach: Iterate over list and transform objects to doubles, 
inserting these doubles into a double[] and passing it to the 
appropriate UnivarStatistic object's evaluate method.

Primarily Issue: The double[] is regenerated for every getXXX call.

Pros: The "change in state" of the external list doesn't need to be 
monitored.

Cons: Not a very efficient means of working with the Lists.

Would it would be better if the state of the list could be monitored and 
the double[] regenerated if an object has been added/removed from the 
list? Are there known means of "listening to Collections".

-Mark


---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org