[Bioc-devel] YAML for specifying, storing and retrieving (many) parameters

[Matthew McCall]

Paul,

YAML seems like a fine option.

If possible, I'd like to submit the ternarynet package to be part of
this effort as well. I'm definitely interested in any effort to
compare / benchmark network inference algorithms.

Best,
Matt

On Wed, Aug 1, 2012 at 1:39 PM, Steve Lianoglou
<mailinglist.honeypot at gmail.com> wrote:
> Hi,
>
> On Wed, Aug 1, 2012 at 1:05 PM, Paul Shannon <pshannon at fhcrc.org> wrote:
>> Here at BioC I am working on making Bioconductor more hospitable to packages which do network inference -- of which we have a few already.  I have learned that such software usually employs many parameters.  How to specify, store and retrieve them?
>>
>> I propose to use YAML (http://en.wikipedia.org/wiki/YAML) as a user-friendly & technically strong solution.  I am writing to solicit feedback on this idea.
>
> I use YAML for the same purpose in a few places -- I only have good
> things to say about it.
>
> -steve
>
>
>
>>
>> Among YAML's virtues are:
>>
>>   - machine readable (lots of language support, including http://cran.r-project.org/web/packages/yaml)
>>   - human readable, easily edited
>>   - represents scalars, lists, associative arrays, multi-line string literals
>>
>> The wikipedia page (http://en.wikipedia.org/wiki/YAML) is a good introduction and has clear, focused examples.
>>
>> At present, only Tony Handstad's submitted 'triform' package uses YAML.  I do not know of other solutions used in BioC.
>>
>> Can any one offer opinions pro or con?  Other alternatives?
>>
>> For any one might be interested, here is a very brief account of what I am working on to support network inference projects, existing and to come.
>>
>> 1) Make experimental data packages of relevant DREAM network inference challenges: (http://www.the-dream-project.org/)
>> 2) Exploring the usefulness of some common input and output classes, in the hope that different inference packages can be easily run, and results compared, for the non-expert user.
>> 3) Identifying common visualization and assessment needs.
>>
>> My first prototype addresses Rich Bonneau's NYU/ISB 'inferelator', which has been around for a few years, has been published  prominently, done well in the DREAM challenges, is written in R, and is famously difficult for outsiders to run.  I hope to learn from, and offer support to Robert Costelo's qpgraph, and Patrick Meyer's minet as well, in addition to anyone else who would like to contribute their ideas or software.
>>
>>  - Paul
>> _______________________________________________
>> Bioc-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/bioc-devel
>
>
>
> --
> Steve Lianoglou
> Graduate Student: Computational Systems Biology
>  | Memorial Sloan-Kettering Cancer Center
>  | Weill Medical College of Cornell University
> Contact Info: http://cbio.mskcc.org/~lianos/contact
>
> _______________________________________________
> Bioc-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/bioc-devel



-- 
Matthew N McCall, PhD
112 Arvine Heights
Rochester, NY 14611
Cell: 202-222-5880