See answers below.
On 07/31/2012 12:00 PM, Gilles Sadowski wrote:
On Tue, Jul 31, 2012 at 08:18:13AM +0200, Dennis Hendriks wrote:
* Why doesn't "KolmogorovSmirnovDistribution" implement one of the
interfaces of the package (and/or inherit from an abstract class)?
See also:
http://apache-commons.680414.n4.nabble.com/math-KolmogorovSmirnovDistribution-not-part-of-the-distribution-hierarchy-td4114395.html
Thanks for digging that thread out.
Do I rightly deduce that it is unfinished work, which should give rise to a
JIRA ticket?
That would be my guess.
* Why is method "probability(double)" part of the "RealDistribution"
interface? [All implementations return 0.]
My guess would be that it is left over from when integer and real
distributions shared a common hierarchy.
I would agree with the guess.
Does everyone agree to deprecate this method (in 3.1) and remove it (in
4.0)?
+1
* Shouldn't the "cumulativeProbability(double x0, double x1)" method be
renamed "probability(double x0, double x1)"?
Does everyone agree to deprecate this method (in 3.1) and remove and replace
it (in 4.0) with a method named "probability(double x0, double x1)"?
I would prefer to:
- add new method (3.1)
- deprecate old method (3.1)
- move impl to new method (3.1)
- redirect old method to new method (3.1)
- remove old method (4.0)
That way, we can now change our code to use the new method, and not get any
deprecation warnings on use of the old deprecated method, for which no
alternative exists...
Regards,
Gilles
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
Dennis
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscr...@commons.apache.org
For additional commands, e-mail: dev-h...@commons.apache.org
