3 ms·
From my experience, one works a lot explorative in REPL terminals (like IPython, R Studio, MATLAB, etc, depending on your field). This is the place where to cre
by ktpsns 8y ago
From my experience, one works a lot explorative in REPL terminals (like IPython, R Studio, MATLAB, etc, depending on your field). This is the place where to create novel graphs and plots and to discuss them with colleagues in order to make conclusions. Discussion also takes place in regular meetings, completely depending on the kind of workplace, of course.
Organizational tasks like preparing a work setups, sorting, downloading or uploading data also takes a significant time. It is always a tradeoff of automating tasks (writing scripts to do the painful work) or doing it manually by hand.
- kmalhi 8y agoIn what scenarios do you have to manually download and upload data. There are ingestion tools out there. It seems like automating tasks is painful. What kind of tasks are those? I’m guessing all of this takes a lot of your time away from actual analysis
- ktpsns 8y agoYup, there are tools "for almost everything", but they must be flexible enough, then compatible to the machines used, and learned to use. For instance, in science, "grid computing" is popular to automate away the orchestration of files and codes between machines. But it's not always available and sometimes not applicable. I assume every workplace has an overhead part which keeps one from doing the actual work.
- kmalhi 8y agoDo you have to do the whole infrastructure setup yourself or do you get help from your engineering team?