DescriptionReplicability of scientific work based on computation is a subject which has been receiving increased scrutiny recently. One approach to replicating a computational finding is to run the original source code. Availability of source code however is not routine; Only 3/33 computationally based article authors released source code from JASA in 2006, and a 2015 study showed that only 44% of computer science article authors released their source code. The field of Computational Physics has yet to be examined in this way, nor has the effect of author knowledge of such a study been measured.
We present our findings regarding the availability of source code in recent articles of the Journal of Computational Physics as well as how author knowledge of the study affects their willingness to make source code available. This work extends current reproducibility efforts being explored by the ACM, SIGHPC, and the SC conference community.