You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@commons.apache.org by Al Chou <ho...@yahoo.com> on 2003/06/20 23:53:42 UTC

Re: [math] Remove mode, remove some methods from StatUtils

--- "Mark R. Diggory" <md...@latte.harvard.edu> wrote:
> bugzilla@apache.org wrote:
> >Also, I neglected
> >to execute maven clean before test.  My mistake.  I see now that these
> methods
> >are reused, so we have to keep them in StatUtils (unless we decide to drop
> >skewness and kurtosis from UnivariateImpl).  I will submit a revised patch
> >keeping these things in; but I am getting a little squemish about so many
> static
> >methods. My preference would be to remove skewness and kurtosis from
> >UnivariateImpl and StatUtils as well.
> >
> >
> Ok, I'm always going to be a little touchy about Skew and Kurt, their 
> kinda like my first born children.

We really should look at whether the functionality we are providing is useful
to our audience and base our decisions heavily on that.  I understand your
sentiment, but what it should boil down to is, "Is this feature providing
value?"  For instance, now that I have backed off on submitting a Ridders'
method root finder, I no longer have an explicit use case for MathUtils.sign(),
and it could be argued that this method should be removed.  As for skewness and
kurtosis, my opinion remains that they are little used in the real world (and
are thus causing undue debate within the project and wasting time and energy). 
No one has commented on that remark, as far as I can remember.


Al

=====
Albert Davidson Chou

    Get answers to Mac questions at http://www.Mac-Mgrs.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


Re: [math] Remove mode, remove some methods from StatUtils

Posted by robert burrell donkin <ro...@blueyonder.co.uk>.
On Saturday, June 21, 2003, at 12:14 AM, Mark R. Diggory wrote:
> Al Chou wrote:

<snip>

>> As for skewness and
>> kurtosis, my opinion remains that they are little used in the real world 
>> (and
>> are thus causing undue debate within the project and wasting time and 
>> energy). No one has commented on that remark, as far as I can remember.

<snip>

> All I can say as a developer on this project is, I use them, I enjoy 
> working on them, (they are moments as natural as mean and variance and 
> they natural incusions into descriptive statistics). I've worked hard on 
> their development, and I'd really like them to stay part of the package.

+1

an approach which has proved useful in the past with unreleased code of 
this kind is to keep features which have been developed in the code base 
as long as possible. it's much easy to make decisions about scope (what's 
in and what isn't) for a particular release in the immediate run up. (the 
only downside about this approach is that is really annoys  authors with 
books in preparation...)

- robert


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


Re: [math] Remove mode, remove some methods from StatUtils

Posted by "Mark R. Diggory" <md...@latte.harvard.edu>.
Al Chou wrote:

>>Ok, I'm always going to be a little touchy about Skew and Kurt, their 
>>kinda like my first born children.
>>    
>>
>
>We really should look at whether the functionality we are providing is useful
>to our audience and base our decisions heavily on that.  I understand your
>sentiment, but what it should boil down to is, "Is this feature providing
>value?"  For instance, now that I have backed off on submitting a Ridders'
>method root finder, I no longer have an explicit use case for MathUtils.sign(),
>and it could be argued that this method should be removed.  
>
[sigh] I feel the debate should be more about where to put things we 
have already worked hard to create. rather than to eliminate them to 
make a release. If you intend to implement Riddlers later after release 
rather than before, it would be wise to keep sign around.

My time and efforts are focused into the features I want to see 
available in the statistics library (and the features I *enjoy* working 
on), others do the same and the library grows in functionality. Be very 
carefull about taking away the things that other developers on the 
project are working with great interest on.

>As for skewness and
>kurtosis, my opinion remains that they are little used in the real world (and
>are thus causing undue debate within the project and wasting time and energy). 
>No one has commented on that remark, as far as I can remember.
>
>
>Al
>  
>
All I can say as a developer on this project is, I use them, I enjoy 
working on them, (they are moments as natural as mean and variance and 
they natural incusions into descriptive statistics). I've worked hard on 
their development, and I'd really like them to stay part of the package.

-Mark

-- 
Mark Diggory
Software Developer
Harvard MIT Data Center
http://www.hmdc.harvard.edu



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