Abstract
This article strives to expand on existing work to demonstrate advancements in data processing made available using event mode measurements. Most spallation neutron sources in the world have data acquisition systems that provide event recording. The new science that is enabled by utilizing event mode has only begun to be explored. In the past, these studies were difficult to perform because histograms forced dealing with either large chunks of time or a large number of files. With event based data collection, data can be explored and rebinned long after the measurement has completed. This article will review some of the principles of event data and how the method opens up new possibilities for in situ measurements, highlighting techniques that can be used to explore changes in the data. We also demonstrate the statistical basis for determining data quality and address the challenge of determining how long to measure mid-measurement. Finally, we demonstrate a model independent method of grouping data via hierarchical clustering methods that can be used to improve calibration, reduction, and data exploration.
| Original language | English |
|---|---|
| Article number | 093001 |
| Journal | Review of Scientific Instruments |
| Volume | 89 |
| Issue number | 9 |
| DOIs | |
| State | Published - Sep 1 2018 |
Funding
The work of D.O. was funded by the BES Early Career Award, Exploiting Small Signatures: Quantifying Nanoscale Structure and Behavior KC04062, under Contract No. DE-AC05-00OR22725. Neutron scattering measurements are sponsored by the User Facilities Division, Office of Basic Energy Sciences, U.S. Department of Energy. Portions of this research used data from the NOMAD, POWGEN, and VISION instruments at the Spallation Neutron Source. The authors would like to thank Ashfia Huq for providing the POWGEN data, Katharine Page for providing the NOMAD data, and John Faber for insights into pattern matching metrics.