On 9/9/19, D.S. McNeil <dsm...@gmail.com> wrote: > [coming over from the pydata post] > > I just checked about ~150KLOC of our Python code in a financial context, > written by about twenty developers over about four years. Almost every > function uses numpy, sometimes directly and sometimes via pandas. > > It seems like these functions were never used anywhere, and the lead dev on > one of the projects responded "never used them; didn't even know they > exist". I knew they existed, but even on the rare occasion I need the > functionality I need better control over the dates, which means for > practical purposes I need something which supports Series natively anyhow. > > As it is, they also clutter up the namespace in unfriendly ways: if there's > going to be a top-level function called np.rate I don't think this is the > one it should be. Admittedly that's more an argument against their current > location. > > Although it wouldn't be useful for us, I could imagine someone finding a > package which provides numpy-compatible versions of the many OpenFormula > (or > whatever the spec is called) functions helpful. Having numpy carry a tiny > subset of them doesn't feel productive. > > +1 for removing them. > > > Doug
Thanks Doug, that's useful feedback. Warren > > > > -- > Sent from: http://numpy-discussion.10968.n7.nabble.com/ > _______________________________________________ > NumPy-Discussion mailing list > NumPy-Discussion@python.org > https://mail.python.org/mailman/listinfo/numpy-discussion > _______________________________________________ NumPy-Discussion mailing list NumPy-Discussion@python.org https://mail.python.org/mailman/listinfo/numpy-discussion
