NCRR Meeting Notes
Wednesday, August 16, 2000, 12:00 pm
Rob MacLeod and Dave Weinstein
With these meetings we hope to achieve the following goals
- Make sure all members of the NCRR group know each other
and are aware of both the long terms goals and the course of
progress of the projects.
- Report progress, present ideas, and
solicit feedback and ideas in front of the group.
- Constantly review and upgrade priorities of the various
projects and ensure that tangential concerns do not distract from
the progress of the group.
- Maintain momentum on the high priority projects.
- Vis 2000 Demos:
- Reminder about the Visualization 2000
conference in October and that we will have an evening visit from
conference attendees to show projects within the School of
Computing. For this, we need to develop some demos that emphasize
the application area of our activities. Please contact Dave or
Rob for specific ideas of what you would like to demonstrate. We
will also be approaching people to help develop and present other
demos.
- Meeting time:
- After marking up the schedule, we came to the
following free meeting times:
Monday 10:00-12:00
Tuesday 9:00-12:00
Wednesday 10:00-12:00
3:00-4:00
- Annual Report:
- Erik has the progress report available on
the web site at
www.sci.utah.edu/ncrr/pubs/progress.html. Please check it
out and report any problems to Erik.
- Sourceforge:
- Marty reported on progress he was making with
this tool and the meeting to discuss its use for managing BioPSE.
Sourceforge is a comprehensive tool for open source type releases
and includes support for CVS, web interface to the source code, bug
tracking and a number of other tools for managing large code
projects.
- Fields class:
- We did not have time to hear Eric's
presentation on the fields class but we are very close to getting a
working version out. Alexi and Michael are also helping Eric to
test the fields class and refine it for general use.
- Flight path module
- Alexi has the flight path working and
will work with Yarden to create some fly through examples. This
module also support generation of images for creating mpeg/QTime
animations that we can use for talk and on the web site. He is
working on a stereo implementation now.
- BEM Module:
- Guo Tang has completed a module for SCIRun that
implements some Fortran source codes for solving field problems
using the boundary element method. He will present an overview of
how he implemented this sort of code wrapping at the next meeting.
- Module review:
- Note that for each of the modules we
review, there should emerge a design document. Ted gave us a nice
example this week so see his spec on the histogram module and the
documentation Section 3.4 for more details on
the layout.
- Histogram:
- we began the module review with a second
look at the histogram module. Ted has prepared an excellent
document about the design of the module (available for viewing
at
www.cvrti.utah.edu/ dustman/histo/histo.html
). We
agreed on the layout described in the document and so that
module is ready to move to implementation.
- Showfield
- Marty had nothing to report on this module but
one outcome of the discussion of the histogram was to have
modules, perhaps part of showfield, for displaying 2D and 1D
fields; these could be generalized to handle things like
histograms but also time signals, bar charts, and any other
display modes that we will need.
- Linear algebra modules
- Leah presented some ideas for
modules to handle basic linear algebra manipulation. After
considerable discussion, we agreed on several points:L
- we would try and treat all matrices and vectors as a basic
matrix datatype like Matlab does; this will require some
interfacing with existing vector and matrix datatypes but has
advantages in the interface and builds on the familiarity
that users have with Matlab and other high level tools for
matrices.
- to pursue a two-path strategy for these modules. For smaller
matrices we would build on the BLAS libraries and potentially
other packages that would exist on the local machine (e.g.,
LAPACK, PETSc). For larger systems we would try and use the
Netsolve
system that we are currently investigating together with
Michelle Miller and the other Tennessee folks. The system
allows for ``remote'' solving of systems where remote might
be actually on the same machine, another machine in a local
cluster, or a computing facility accessed via the Internet.
- we agreed on a basic set of operators we would
implement as methods and allow as entries in the UI for this
module: (), +, * (parentheses for grouping, addition,
subtraction, multiplication)
- Volume viewer
- Chris M. reported on a module to view
volumetric data with separate viewers for each of the 1D, 2D,
and 3D cases.
- Module list:
- Dave is still working on an updated list of modules
that we will review next week and hopefully finalize (with some
priority scores) and thus establish the goals for release 1.0 of
BioPSE.
- Upcoming conferences:
- Conferences coming up that are of special
relevance to the group include:
- Biomedical Engineering Society (BMES): October 12-14, 2000
in Seattle (Rob will be chairing a session).
www.engr.washington.edu/~uw-epp/bmes/
- Visualization 2000: Oct. 09, 2000 - Oct. 13, 2000, Salt
Lake City.
- 7th International Conference on
Numerical Grid Generation in Computational Field
Simulations: September 25-28, 2000, Whistler, British
Columbia, Canada. We would like to send Xinlong to this,
but are not sure if he will be here in time. Dave and Rob
are both busy and may not be able to attend but we would
like to have someone attend and report back.
- NFSI III: September, 2001, Innsbruck, Austria
- NCRR Seminars
-
- Tom Feree from U of Oregon will be coming September 3
to spend some days here working with Dave and Leah and
will also give an NCRR seminar. Title and abstract are in
the works and will appear on the web site shortly
- Martin Schweiger from Simon Arridge's lab in London,
England will once again visit us and we will ask him to
present an NCRR seminar. This project involves diffuse
optical tomography. Chris J. will be the contact on this
visit.
Documentation
- We discussed the mechanisms by which we can and should
produce the design docs (examples include Eric's and Ted's). The
advantage of LaTeX is that ease with which many of us can create
documents and see them in printed form quickly, and then convert to
HTML using latex2html. The disadvantage is that this is not part
of the documentation build and hence requires subsequent conversion
to the XML format. Another option is to enter the information
directly into an XML file and then generate everything from that.
In order to implement the latter option, we need to have a template
file available and also some tools to edit and view the results.
Marty has some of this on his machine and will try and get it into
a form we can all use. He will also investigate the changes we
need in the apache server to allow direct viewing of XML from a web
browser.
- Erik and Rob are still working together on ways to convert LATEX
to HTML in a consistent layout and format for inclusion on the web
site with minimum work.
- Ted's overview of the document layout for the PSE is at
www.sci.utah.edu/scirun_docs/doc/docMap.html
- Erik has created some nice style sheets for the XML based
module documentation system.
- Map3d progress:
- Rob has ported map3d to the CVRTI and is
now compiling a list of additions and features to try and address
before the release date.
- CVS for Map3d
- Chris M., Ted, and Rob met to discuss where
to keep the source code for map3d and came up with a scheme whereby
the general purpose libraries that map3d uses should live on the
CVRTI CVS server but the main map3d code should stay in the SCI
server. We will try and implement this scheme and see if it works.
- Release schedule:
- the plan is to have map3d released by
September 1, 2000.
- Documentation:
- there already exists a manual for map3d but this
will need updating prior to release. Rob will update the manual to
reflect changes in the program and well as reduce the options to
match the released version.
- Simulation post doc position:
- We are still discussion the
excellent candidates we have for the inverse problems
post-doc/staff scientist position. Please send comments to Chris
if you have any input.
- Xinlong Wang:
- will be coming to Utah as the meshing post-doc and we
expect him sometime in September or early October.
- Laura's project :
- due back from vacation this week.
- Chris Butson's project:
- no news
- Leah's project:
- Leah is working with Dave to create a module
that will perform impedance tomography analysis of some simple
models using the finite element formulation that is similar to that
which we use in forward problems.
- Rob v. Uitert:
- Rob is starting to work with Dave and Sri
Nagarajan from Bioengineering on a possible project. More
details to follow.
- Data archive:
- No progress on this. But more requests for data
have arrived.
- Tracking hits:
- No progress on this.
- Web site parsing:
- No news on this
- BioPSE logo:
- Erik is working with Rob on a logo for BioPSE.
- NCRR Poster:
- The poster is completed and on its way. Erik will
put a link to it on the web page.
Agenda for for future meetings
Specific agenda items for the next meeting include:
- Update on NCRR seminar speakers and scheduling their talks
- Schedule for NCRR meetings
- Guo Tang presentation of BEM module
- Marty's presentation of the XML module template and tools for
editing and viewing them
- Eric's presentation on Field Class
- Dave's presentation and discussion of complete module list for
BioPSE version 1.0.
Most of the next meetings will be organized around the brief reports from
all present and then discussion of modules. Each person in the group has
volunteered for one more more modules and will lead the discussion, which
will center on the design and capabilities of the module. For details on
the overall layout of modules and a list of BioPSE modules we have
envisioned so far, go to
www.cvrti.utah.edu/~macleod/ncrr/technical/modules-19jul00.shtml
NCRR Meeting Notes
Wednesday, August 16, 2000, 12:00 pm
This document was generated using the
LaTeX2HTML translator Version 99.2beta6 (1.42)
Copyright © 1993, 1994, 1995, 1996,
Nikos Drakos,
Computer Based Learning Unit, University of Leeds.
Copyright © 1997, 1998, 1999,
Ross Moore,
Mathematics Department, Macquarie University, Sydney.
The command line arguments were:
latex2html -split 3 -no_white -link 3 -no_navigation -nomath -html_version 3.2,math notes-16aug00
The translation was initiated by Rob MacLeod on 2000-08-19
Rob MacLeod
2000-08-19