Calvera: A Platform for the Interpretation and Analysis of Neutron Scattering Data

Data analysis for neutron scattering experiments is driven by the scientific needs of the instrument users and varies greatly by technique and field of study. Data from an experiment must first be “reduced” so that instrument artifacts are removed, and then scientists must choose from a wide variety of tools and applications to assemble a workflow that enables useful scientific results to be extracted. The highly manual nature of this process, combined with difficulty accessing computational resources and data when needed, puts limits on the efficiency and nature of the analysis undertaken. In addition, other activities, such as tracking data provenance to ensure the analysis is reproducible, or providing live data analysis during experiment runs, are also difficult to achieve. Calvera is a platform that aims to solve many of the difficulties encountered by scientists as they analyze experimental neutron scattering data. In particular, the platform will provide an integration point for a range of services, such as data virtualization, remote computation, and visualization under the control of a workflow management system. In addition, the platform will handle security related issues, and maintain a history of the data sets employed during workflow execution. User’s will be able to construct, manage, and share workflows via a graphical user interface, as well as script workflows via a python API. In this paper, we will describe the architecture and design of Calvera, as well as how we will address the many requirements for executing neutron science workflows in a distributed environment.