ORSO workshop 2020

Session 1: Detailed introduction and aims of the working group – (26th 18:00/27th 08:00)

  • Selection of a scribe
  • Introduction to current members of group
  • The goal of this group is to share as much as possible of the code and practices that have been developed worldwide for analyzing reflectivity data
  • This primarily means modeling and fitting - issues related to data reduction (standardisation of format, and reproducibility) have been sorted into the other working groups

  • Tasks for the workgroup:

    • Verifying that the various engines that perform reflectivity calculations give the same results for the same inputs (validation)
    • Defining a shared vocabulary for describing reflectivity model problems (that can be used to verify the calculation engines)
    • Defining “dialects” for domain-specific applications (e.g. magnetic layered structures, membrane protein systems, polymer melts, etc)

  • Some work (by A. Nelson) already done on verification task

  • Discussion topics:
    • Existing tasks: prioritize
    • Existing tasks: identify “next step” (action item)
    • Other tasks: what else can we do to share data analysis between reflectometrists?

Desired output from this session:

  1. Priorities for tasks
  2. New tasks
  3. Action item(s) for tasks

Session 2: Toward a shared calculation engine – (26th 22:30/27th 18:00)

…or at least a shared language for the inputs to the engine

  • Selection of a scribe
  • What is the best way to collaborate on a shared reflectometry calculation engine?
    • Just the kernel (written in C or ?)
    • A higher-level program that takes whole problem definitions (parameters, constraints, etc)
  • What are the required elements for describing the inputs to a reflectivity calculation kernel?
  • Build on the work already done on validation (a programmatic adapter for each engine being hand-crafted by A. Nelson)

Desired output from this session:

  1. A strategy for collaborating on reflectometry calculator
  2. Discussion on standardisation of problem language
  3. Any new items

Session 3: Specific problem definitions: building real models – (27th 22:30/28th 08:00)

Dealing with distinct reflectometry problem categories, e.g. magnetic layers, polymer melts, biological membranes…

  • Selection of a scribe
  • Problem categories might naturally follow implementations of fitting routines (everyone with a very specific fitting problem has had to address this somehow), so…
  • How many fitting programs are out there? (some may support more than one type of model / problem definition)
  • Your own fitting problem: What terms would you need to describe the fits you do? Constraints?
  • How can we collaborate on constructing model builders for all these categories of problem?
  • Currently calculation/fitting engine and model builder are all one application: how to effectively share calculation engine between model builders

Desired output from this session:

  1. Identification of problem domains (dialects) people are working on
  2. Strategy for collaboration on model builders
  3. Strategy for sharing calculation engine between model types (outputs of model builders)
  4. Any new ideas for collaboration on domain-specific fitting