commons-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Luc Maisonobe <Luc.Maison...@free.fr>
Subject Re: [math] improvement of the ODE package
Date Sun, 13 Jul 2008 09:57:31 GMT
Phil Steitz a écrit :
> Luc Maisonobe wrote:
>> In order to solve issue MATH-172 and for other related needs, I have
>> started work on the ordinary differential equations package as some of
>> you may have noticed.
>>
>> The first phase was simple cleaning: reorganizing the package that was
>> becoming quite huge. There are now several subpackages:
>>  - nonstiff which provides integrators (and continuous output
>>    interpolators) for non-stiff problems
>>  - sampling which deals with step handlers (fixed and variable steps)
>>  - events which deals with discrete events (what used to be called
>>    switching functions in the previous version)
>>
>> The second phase is under development: introducing new simple multistep
>> integrators: fixed step explicit Adams-Bashforth, fixed step implicit
>> Adams-Moulton and BDF, still for non-stiff problems. This is a little
>> harder than I thought at first, because I really want this integrators
>> to provide the same rich features as the existing ones: multiple
>> discrete events handling, continuous output, events handlers. I also
>> wanted a user customizable autostart feature (multistep methods need
>> some help to get started). I did not find any reference to continuous
>> output models for these simple integrators, so I had to develop the
>> underlying equations myself. I hope to check a few new classes soon.
>> Meanwhile, I also have to change a little the shared API with the
>> existing integrators, in order to get a consistent package.
>>
>> The third phase will consist in implementing integrators able to handle
>> stiff problems. I will probably start with a port of LSODE (from netlib,
>> available under a BSD type license) which is based on GEAR method.
>>
>> Luc
>>
>>   
> Just got this
> Test set: org.apache.commons.math.ode.ContinuousOutputModelTest
> -------------------------------------------------------------------------------
> 
> Tests run: 4, Failures: 1, Errors: 0, Skipped: 0, Time elapsed: 0.053
> sec <<< FAILURE!
> testModelsMerging(org.apache.commons.math.ode.ContinuousOutputModelTest) 
> Time elapsed: 0.018 sec  <<< FAILURE!
> junit.framework.AssertionFailedError: expected:<1.0> but was:<NaN>
>        at junit.framework.Assert.fail(Assert.java:47)
>        at junit.framework.Assert.failNotEquals(Assert.java:277)
>        at junit.framework.Assert.assertEquals(Assert.java:98)
>        at junit.framework.Assert.assertEquals(Assert.java:105)
>        at
> org.apache.commons.math.ode.ContinuousOutputModelTest.testModelsMerging(ContinuousOutputModelTest.java:129)

Ooops. Sorry for that. I hope this is fixed now.

Thanks for reporting it.

Luc

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



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


Mime
View raw message