Let me clarify. I do mean to suggest that "developers" use Jupyter environment, including Notebook, for developing production quality Python code. Most enterprises don't use Python for web development and I don't what developers use. I do know that *researchers* in scientific computing and data science who use Python use Jupyter extensively, not just for "tutorials" and "demos". I don't have data on this. It's anecdote. The IDE for researchers using R is, of course, R Studio.
Apache Zepellin and notebook environment is also worth looking at because of its Spark integration. https://zeppelin.incubator.apache.org/ On Sun, Apr 3, 2016 at 2:04 PM, Matthew Brett <[email protected]> wrote: > On Sun, Apr 3, 2016 at 10:46 AM, Matthew Gidden > <[email protected]> wrote: > > Hi all, > > > > I agree. I spend probably 70% of my time in notebooks. However, the > > distinction is between "initial coding" vs. "coding". In my specific > > environment (which I do not claim is authoritative or representative), I > > have found many users stay in the IDE/notebook environment. > > I've seen this too. > > I'd be very interested if there is any data about the effect this has > on errors in the resulting analysis, and code quality. > > The notebook is an excellent tool for writing tutorials and demos, but > I have almost stopped using it for my work (which is a combination of > development, teaching and research). I found the notebook GUI was a > serious barrier to stepping back and thinking hard about the problem. > It encouraged me to play with things until they worked, which is a > temptation I have to work very hard to avoid. > > Best, > > Matthew >
_______________________________________________ Discuss mailing list [email protected] http://lists.software-carpentry.org/mailman/listinfo/discuss_lists.software-carpentry.org
