[SciPy-user] Bayes net question
Karl Young
Karl.Young at ucsf.edu
Thu Feb 7 15:19:27 EST 2008
Great, sounds to me like the first question we need to answer then
(should be easy now) is whether the approach should be to just start
porting BNT functionality into the OpenBayes framework, or to first
review the OpenBayes framework in some detail re. whether anyone thinks
we should do any refactoring.
>OpenBayes was an earlier attempt to port BNT to python. It uses similar
>network representation.
>
>- Ed
>
>On Thu, 7 Feb 2008 10:26:05 -0800
>"Young, Karl" <karl.young at ucsf.edu> wrote:
>
>
>
>>Sounds good, thanks for the update. Given the conversations subsequent to this post re. establishing that Kevin Murphy has given the go ahead re. porting and relicensing his Bayes Net matlab toolbox for integration into scipy, it sounds like one of the first orders of business for those who signed up for this project might be to try to establish whether the Openbayes project, or minor modifications/extensions, could provide a useful framework for the port. I'll continue to look at both but my understanding at this point is that the matlab toolbox currently has significantly more functionality than the Openbayes project. Looking at Kevin Murphy's description of his toolbox it seems he took a broader approach than some Bayes net packages re. general approaches to estimation in graphical models (which was attractive to me) and that might entail a different approach to the structure of the package than the author of Openbayes took (but this is just speculation at this point - I
>>
>>
>
> 'l
>
>
>> l shut up).
>>
>>Karl Young
>>Center for Imaging of Neurodegenerative Disease, UCSF
>>VA Medical Center, MRS Unit (114M)
>>Phone: (415) 221-4810 x3114
>>FAX: (415) 668-2864
>>Email: karl young at ucsf edu
>>
>>
>>
>>-----Original Message-----
>>From: scipy-user-bounces at scipy.org on behalf of Ed Rahn
>>Sent: Thu 2/7/2008 9:35 AM
>>To: scipy-user at scipy.org
>>Subject: Re: [SciPy-user] Bayes net question
>>
>>The author of Openbayes does not mind integrating it into scipy, the
>>discussion can be found in the attached email.
>>
>>>From this repo
>>http://svn.berlios.de/svnroot/repos/pybayes/branches/Public
>>I have converted it from numarray to numpy, the patch can be found at:
>>http://lamedomain.net/openbayes/numpy.diff
>>
>>- Ed
>>
>>On Mon, 14 Jan 2008 14:45:49 -0800
>>Karl Young <Karl.Young at ucsf.edu> wrote:
>>
>>
>>
>>>I'm starting to play with Bayes nets in a way that will require a little
>>>more than just using some of the black box packages around (e.g. I'd
>>>like to play around with using various regression models at the nodes)
>>>and would love to do my exploring in the context of SciPy but I didn't
>>>see any such packages currently available. I did find a python package
>>>called OpenBayes (http://www.openbayes.org/) that after a very cursory
>>>examination looked pretty nice but apparently is no longer being
>>>developed. Does anyone know if there has ever been any discussion with
>>>the author of that package re. incorporating it into SciPy ?
>>>
>>>--
>>>
>>>Karl Young
>>>Center for Imaging of Neurodegenerative Diseases, UCSF
>>>VA Medical Center (114M) Phone: (415) 221-4810 x3114 lab
>>>4150 Clement Street FAX: (415) 668-2864
>>>San Francisco, CA 94121 Email: karl young at ucsf edu
>>>
>>>_______________________________________________
>>>SciPy-user mailing list
>>>SciPy-user at scipy.org
>>>http://projects.scipy.org/mailman/listinfo/scipy-user
>>>
>>>
>>>
>>
>>_______________________________________________
>>SciPy-user mailing list
>>SciPy-user at scipy.org
>>http://projects.scipy.org/mailman/listinfo/scipy-user
>>
>>
>>
>_______________________________________________
>SciPy-user mailing list
>SciPy-user at scipy.org
>http://projects.scipy.org/mailman/listinfo/scipy-user
>
>
>
--
Karl Young
Center for Imaging of Neurodegenerative Diseases, UCSF
VA Medical Center (114M) Phone: (415) 221-4810 x3114 lab
4150 Clement Street FAX: (415) 668-2864
San Francisco, CA 94121 Email: karl young at ucsf edu
More information about the SciPy-User
mailing list