Reliability
Notes from Breakouts
The comments from breakout groups discussing reliability and reproducibility are provided below:
The previous work from the Working group has been documented on the wiki at http://www.cansas.org/wgwiki/index.php/Reproducibility_and_Reliability_Working_Group
Past messages for the mail group are available at: https://groups.google.com/forum/#!forum/cansas-rrwg and there is a button marked 'Apply to join group'.
Summary from Plenary
During the wrap up the primary topic that came up in many different ways from several groups was the idea of developing new standards in some way and doing round-robins on those. Some discussion of how to achieve that was had during the short plenary wrap up in fact. Other topics that came to the plenary included including multiple scattering warnings and modeling into data reduction and analysis software and documenting good practices in data reduction. finally, two topics brought up from the individual breakouts: multi mode methods and analysis; and database/repository of careful measurements with appropriate metadata were deemed to be topics that we be presented and discussed in day 2: data analysis.
Agreed actions from the meeting on 5 June 2017 include:
- Prepare some new potential Q, I, and resolution standard materials, covering a broad range of measurements including soft X-rays and GI techniques. Then use in round-robin to test utility.
- A group at NIST (Joe Kline, Ron Jones, Paul Butler) will meet to discuss what type of samples (focusing probably on 'grating-type') could be prepared. Nick Terril from Diamond will participate by telecon.
- Chris Tassone will send out an email to the list of participants asking for willing to participate in measurements of new round-robin samples.
- Work to include multiple scattering in reduction and analysis codes
- Working group volunteers: Grethe Jensen, Tim Snow, Brian Pauw, Javier Perez, Tom Caswell?
- Prepare a guide on best practices in data reduction for a few standard classes of experiments.
- Brian Pauw leading: a number of people wanted to join will email Brian their interest so he can capture the names.
