DANSE Diffraction Software for the SNS: DiffDANSE

1
DANSE Diffraction Software for the SNS DiffDANSE

• S.J.L. Billinge
• Dept. Physics and Astronomy
• Michigan State University

2
Outline

• Local Goals of DiffDANSE
• Proposed scope and tasks
• Preliminary work
• Communications

3
DiffDANSE goals

• Make diffraction data analysis software that SNS
  users (and other people) want to use!
• Make sure that it has a long life

4
DiffDANSE goals

• Make diffraction data analysis software that
  people want to use!
• Solves peoples scientific problems, opens new
  scientific horizons
• Shortens time to publication
• Flexible
• Easy to use
• Good user documentation
• Robust and Stable

5
DiffDANSE goals

• Make sure that it has a long life
• Maintainable (esp. by others, i.e., SNS!)
• Good developer documentation
• Requirements/use cases/tests stored with the code
• Extensible
• modular
• Scalable
• Dynamic memory allocation
• Distributable/parallelizable
• Quality controlled
• Open source

6
Scope
7
Common tasks
Libraries/common utilities
Education/outreach
8
DiffDANSE development team
?
9

• Prototype Dream-desktop for local-structure
  studies using the atomic pair distribution
  function (PDF)

10
Prototype application PDFgui

• Goals
• PDF is a great technique for studying local
  structure, but our programs are clunky and hard
  to useok for hard-core physicists but not for
  wider adoption. Give it an easy-to-use interface
• Often we want to refine multiple datasets as a
  function of temperature/pressure/composition,
  etc.. Make a program that is straightforward to
  automate for multiple datasets but flexible
  enough to give the user control over this process
• Users are interested in structure, not scattering
  data. Allow them to interact with the structure
  straightforwardly in the same program in real
  time
• Make it sufficiently fast that people want to use
  it
• Use it as a prototype to learn how the DANSE
  software development process is going to go

11
PDFgui

• Solution
• Write Python bindings into our existing PDFFIT2
  C code to access variables at the python layer
  at run-time
• Wrap an existing plotting program, matplotlib,
  for 1D plots which support user interactions
  (zoom, explore, etc.)
• Wrap an existing structure visualization program,
  ATOMEYE, for simple structure visualization
• Build a GUI front-end
• We want real-time interactivity so
• Graphics can eat up system resources gt build in
  distributed functionality. Have PDFFIT2 running
  on a remote computer but keep the real-time
  interactivity

12
PDFgui component structure
Plot Control
Matplotlib
StruVis Control
Pymol
13

• Configuring a set of fits for a T-series on
  LaMnO3 data.

14
Plotting pane of PDFgui
15
PDFgui

• Make code people want to use
• Careful design of the structure of the GUI. This
  went through several iterations and ultimately a
  complete refactoring as we struggled with how
  people will use it for different tasks.
• User input
• Rapid development cycle was just a few months
  from inception to the alpha-0 release (local
  release with a single installation). Now working
  on alpha-1 (wider release, some feature
  improvements)
• (already expert PDF) Users were up to speed after
  one tutorial session with no documentation (will
  address at beta release)
• Software control on TRAC and svn allows users to
  post bug-reports and feature requests that wont
  get lost.

16
PDFgui

• Make sure that it has a long life
• Modular
• Clear separation of gui from program logic.
  Interaction between the program and the gui
  through proxies
• Team development results in code that is readable
  by more than one human being
• Coding standards for docstrings and basic syntax
  structure
• Project under subversion control under TRAC

17
(No Transcript)
18
Summary

• Make codes people want to use
• Open Source
• Community engagement what do you want to have?
• Community involvement what do you want to do?
• Now beginning to develop the detailed software
  requirements. Please tell us what you need!