Discoverable, Accessible, Documented, Reusable, Interoperable: the Imperative to Publish Code in Modern Science
Abstract
Computation has become ubiquitous across all of science. It permits scientists to solve problems that are otherwise intractable for human intuition and carry out sophisticated experiments that would be impossible in real world settings. For modern science to advance, it is essential that the algorithms that underlie scientific computation be accessible in the same way as other scientific practice. Yet, our analysis of a half-century of scientific literature shows that this is not the case. For only a small fraction of papers on research supported by computational modeling is the code publicly accessible.
At CoMSES Net, we are trying to change scientific practice in this regard. The NSF-supported Network for Computational Modeling in Social and Ecological Sciences is one of the few organizations that provide open-access publishing of scientific code. The CoMSES Net Model Library ensures that scientific code is discoverable, accessible, citable, and documented. By treating code as published documents—assigning standardized citations and offering optional peer review and DOIs—CoMSES Net seeks to embed code sharing in normal professional scientific practice. CoMSES Net is now collaborating with other organizations to develop community standards and best practices that can make scientific code more reusable, interoperable, and replicable. Our vision is to enable a new transdisciplinary, integrative, computationally-enhanced science that can help humanity better understand and manage the complex world in which we live.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2018
- Bibcode:
- 2018AGUFMIN43C0906B
- Keywords:
-
- 0520 Data analysis: algorithms and implementation;
- COMPUTATIONAL GEOPHYSICSDE: 1904 Community standards;
- INFORMATICSDE: 1976 Software tools and services;
- INFORMATICSDE: 1978 Software re-use;
- INFORMATICS