What are the main tools and computational frameworks used in the Ecological sciences. I have an interest in building “executable papers” and have been looking at iPython Notebook as a toolset that might enable us to do this quickly. Is Python widely used? Are workflow tools useful in this space?
OpenPub
A collection of thoughts and links from the NCEAS Future of Publishing Working Group
In terms of computational tools, R is starting to become quite big – which is starting to lead people to begin to use LaTeX with sweave or knitr for ‘executable papers,’ as it were. Although that progress is slow.
SAS and MATLAB are also large, although many ecologists (particulary older ecologists) are still very tied to point-and-click analytic software like JMP and the like.
Python is seeing some use, as well as PERL, particularly for Ecologists who interact with Evolutionary Biologists. Those are more common there.
I know only a handful of ecologists that use Python, primarily because R is the tool with the most packages for standard ecological analyses. I personally use and advocate Python—I use it for anything that’s not performance-critical (most analysis, text file processing, etc.). This system looks pretty cool, particularly if it’s fairly smooth to include code written in non-Python languages. I’ll definitely give it a spin.
A tangent: the Julia language is new but looks promising. It has much more of a Matlab flavor, but it has a big advantage of being *fast*. It’s compiled (“just-in-time” via the LLVM compiler infrastructure); the authors say that cloud-based operation and visualization is one of their main goals. I wouldn’t be surprised if an open-source startup company emerges around it.
http://llvm.org/
http://julialang.org/
Also, I had a conversation on Facebook last week on the topic that we don’t have any good software for writing papers. LaTeX is good for a small subset of things and a small subset of people but lacks visual interaction and real-time feedback. Word is mediocre for everyone and everything, and isn’t really suitable for structured document editing. And things like this IPython Notebook, Mathematica, etc. are too narrow to extend to most publications.
I would love to have an elegant, structured document editor that integrates with visualization and analysis tools as the perfect complement to better publishing infrastructure. Too bad it’s, as Cameron said, a zero billion dollar industry right now (unless it, you know, could also replace Word).