Re: science-imageanalysis (Was: [RFC] New task: science-dataacquisition)
Andreas Tille <tillea@rki.de> writes:
> On Thu, 11 Dec 2008, Thibaut wrote:
>
> > I believe Yorick should be there too.
> > http://packages.debian.org/sid/yorick
> >
> > Rationale: it's an interpreted language à la IDL/GDL, many people
> > use this for analysing data, including images (e.g. astronomical).
It sounds like I should add yoric to my list of Numerical Computation
(Matlab/IDL like) software at DebianScience/Physics - what do you think?
>
> >From the description (apt-cache show yorick) I'm unable to see a very specific
> connection to image analysis.
>
> > By the way, a "data analysis" task may make more sense than "image
> > analysis". Depends on how fine grained you want these tasks to be.
>
> IMHO yorick would be a nice target for a dataanalysis task - any opinions
> about such a task and further tragets?
I think that there needs to be a distinction between this task and the
"Viewing" task to avoid having two lists that are almost but not
entirely the same. Exactly what the split should be, I'm not
sure. Perhaps something along the lines of "Viewing - visualising
data, only limited calculation facilities", "dataanalysis - more
sophisticated calculation and modelling".
The list of "Numerical Computation (MATLAB/IDL like)" packages, on the
DebianScience/Physics wiki would be good candidates[1].
octave
matplotlib + ipython + numpy
pdl
scilab
freemat
gnudatalanguage
(and presumably yoric should be added to this list)
Perhaps also the DebianScience/Physics list of Computer Algebra
Systems (for doing symbolic maths) - namely:
Axiom [2]
maxima
sympy
yacas
> Anybody volunteering to edit the tasks files accordingly.
Will just adding the tasks file add it to the list, or is there an
index that needs editing as well?
> I promised to
> kickstart Debian Science tasks files and I hope nobody thinks I failed in
> this - but from a pedagogical point of view I should stop doing this simple
> editing here and leave it to somebody who would focus his work more on Debian
> Science than me (who is busy enough with Debian Med).
Given rough consensus on a task's contents, I can probably do that.
Chris
[1] My objection to including gnudatalanguage in the dataacquisition
task was partially a knee jerk reaction to including a subset of 6
(now 7) packages in every task.
[2] Axiom has a FTBFS bug on sparc that is RC critical at the moment.
Reply to: