You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@commons.apache.org by "Mark R. Diggory" <md...@latte.harvard.edu> on 2003/06/18 17:41:07 UTC
Re: [math] StatUtilsTest (was: cvs commit: jakarta-commons-sandbox/math/src/java/org/apache/commons/math/stat
StatUtils.java)
I'm starting to feel replying to commit messages probibly isn't the best
route to posting on the list. I recommend we not do it or at least
rewrite the subject headings to include the [math] string.
Back to the subject, Yes, your quite right, I feel there are two avenues
of testing going on here:
(1) Direct testing: Copy the StoreUnivariateImplTest and rewrite it to
work with StatUtils as StatUtilsTest..
(2) Indirect testing: All UnivariateTests currently implemented will
test against the StatUtil once those implementations have been altered
to delegate to StatUtils where ever possible. This is currently the case
with UnivariateImplTest which now really testing against StatUtils in
the windowed case (This is what I've been testing against during the
intial dev of StatUtils..
I'm for both approaches as the later won't neccessarily be all inclusive
of StatUtils capabilities.
-Mark
Phil Steitz wrote:
>We need to get test cases committed for StatUtils before adding/modififying
>anything else. I am OK with what has been a added; but we need to comply with
>our stated guidelines, which require test cases for all committed code.
>StatUtils needs its own test cases.
>
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org
Re: [math] StatUtilsTest
Posted by "Mark R. Diggory" <md...@latte.harvard.edu>.
Mark R. Diggory wrote:
> (1) Direct testing: Copy the StoreUnivariateImplTest and rewrite it to
> work with StatUtils as StatUtilsTest..
Done
---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org
Re: [math] StatUtilsTest (was: cvs commit: jakarta-commons-sandbox/math/src/java/org/apache/commons/math/stat
StatUtils.java)
Posted by "Mark R. Diggory" <md...@latte.harvard.edu>.
Phil Steitz wrote:
>Each class should include direct test cases and each new commit should be
>accompanied by test cases. Unless we want to modify the guidelines, we need to
>comply with them. You will notice that we have lots of other classes (e.g.
>RandomData, all of the distributions stuff) that are used by other things, but
>all include direct test cases. The basic principle is very simple -- every
>method needs to be tested.
>
>Phil
>
*Test Infected* :-)
-M.
---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org
Re: [math] StatUtilsTest (was: cvs commit: jakarta-commons-sandbox/math/src/java/org/apache/commons/math/stat StatUtils.java)
Posted by Phil Steitz <st...@yahoo.com>.
--- "Mark R. Diggory" <md...@latte.harvard.edu> wrote:
> I'm starting to feel replying to commit messages probibly isn't the best
> route to posting on the list. I recommend we not do it or at least
> rewrite the subject headings to include the [math] string.
>
> Back to the subject, Yes, your quite right, I feel there are two avenues
> of testing going on here:
>
> (1) Direct testing: Copy the StoreUnivariateImplTest and rewrite it to
> work with StatUtils as StatUtilsTest..
>
> (2) Indirect testing: All UnivariateTests currently implemented will
> test against the StatUtil once those implementations have been altered
> to delegate to StatUtils where ever possible. This is currently the case
> with UnivariateImplTest which now really testing against StatUtils in
> the windowed case (This is what I've been testing against during the
> intial dev of StatUtils..
>
> I'm for both approaches as the later won't neccessarily be all inclusive
> of StatUtils capabilities.
Each class should include direct test cases and each new commit should be
accompanied by test cases. Unless we want to modify the guidelines, we need to
comply with them. You will notice that we have lots of other classes (e.g.
RandomData, all of the distributions stuff) that are used by other things, but
all include direct test cases. The basic principle is very simple -- every
method needs to be tested.
Phil
>
> -Mark
>
>
> Phil Steitz wrote:
>
> >We need to get test cases committed for StatUtils before adding/modififying
> >anything else. I am OK with what has been a added; but we need to comply
> with
> >our stated guidelines, which require test cases for all committed code.
> >StatUtils needs its own test cases.
> >
> >
> >
>
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
> For additional commands, e-mail: commons-dev-help@jakarta.apache.org
>
__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com
---------------------------------------------------------------------
To unsubscribe, e-mail: commons-dev-unsubscribe@jakarta.apache.org
For additional commands, e-mail: commons-dev-help@jakarta.apache.org