Reproducible Computational Science

Computational science has enable significant breakthroughs in science and engineering but, a number of factors make this work dificulty to reproduce. Reproducibility is a foundation of the scientific method so we must to redouble our efforts to make this work easy to reproduce.

Papers

• Social, Behavioral, and Economic Sciences Perspectives on Robust and Reliable Science, Report of the Subcommittee on Replicability in Science Advisory Committee to the National Science Foundation Directorate for Social, Behavioral, and Economic Sciences

Topics

Topic	Presentation	Discussion	Action
Professional Societies	TBD	ACM, IEEE, Other	Move societies and their key conferences towards delivering for free the full package of scientific discovery so the work can be easily reproduced.
Vocabulary	TBD	NSF	Remove the ambiguity in terminology e.g. Reproducibility, Replicability
Conferences	TBD	SC, SIAM, CLSAC	Support conferences in developing best known methods for review and documentation.
Tools	TBD	Notebooks, Runtimes, Portability, Productivity, reproducibility.	Document best tools to inform communities and support the development of tools that are missing
Analytics	TBD	Support analytics communities e.g. combustion, CFD, Astrophysics, etc.	Raise awareness of what is possible

Events

Date	Event	Location
March 8-10th, 2017	National Academy of Sciences Sackler Colloquium Reproducibility of Research: Issues and Proposed Remedies	Washington D.C
February 25 – 26, 2017	Platform SuperCluster Workshop	Beckman Center of the National Academies of Sciences & Engineering University of California at Irvine

Tools

• Project Jupyter
• binder
• everware
• Occam
• Popper Project
• Taverna Workflow Management System
• myExperiment with a good example of use here
• OnlineHPC
• Kepler Project
• VisTrails
• Madagascar
• EasyBuilder
• Spack
• Shifter

Speakers

• Bruce Childers
• Todd Harris
• Jeremy Freeman
• Ward Cummingham and his Federated wiki
• Bernard Rous