I enjoyed working w/ R and R studio to have the code and published pdf always under version control and found that the reproducible research community added readers and feedback and interest from outside my narrowly-defined subject area. I didn't enjoy that frequent package updates would break everything. What kind of coding eco-system does your work exist in?
Ask HN: Should I publish my research code?
361–363 of 363 posts
Re: Ask HN: Should I publish my research code?
#362The Carpentries[0] provide some great resources and training for academics who are interested in this. Check out Software Carpentry[1] and Data Carpentry[2] in particular.
Publishing research code is admirable, and in an ideal world everybody would publish their code and data. That said, we shouldn't pretend that there aren't tradeoffs. Time spent polishing your code to make it presentable is time not spent on other aspects of your research. Time spent developing software development skills is time that could be spent learning new research techniques (or whatever). Reproducible research is great, but it's certainly possible to take it too far at the expense of your productivity/career.
You should also take your own personality into account. If you're a perfectionist you might struggle to let yourself publish research-quality code rather than production-quality code and consequently over-allocate the time you spend prettying up your code.
Re: Ask HN: Should I publish my research code?
#363Two times I have published my research code - both times I have found many other papers/projects plagiarized my work without giving me any credit. This happens way more than you would think, especially if you are working under less known advisor, and at less known university. As the other comment said, if you care about "advancing the science", and won't mind stuff like the above happening, then go for it. In my expe…
This in my experience as well. How did you find it out though?