[Numpy-discussion] #734: interactive docstring search (lookfor)

Hoyt Koepke hoytak@gmail....
Thu Apr 10 16:54:48 CDT 2008

My personal 2 cents is that we should put it in the most easy-to-find
place.  I wish I had the propsed lookfor function at my fingertips
when I started to learn numpy/scipy a year ago, coming from matlab.
Like many others (I assume), I just jumped in and learned what numpy
could do by trial and error and I never noticed numpy.lib.utils.info.
Thus my personal vote (though I understand this may not be the best
and there are arguments against it) is to put a lookfor or info
function in the top level namespace, with an obvious name, so newbies
could easily run across it and have a more enjoyable numpy learning

That's just my humble opinion ...


On Thu, Apr 10, 2008 at 2:42 PM, Pauli Virtanen <pav@iki.fi> wrote:
> Travis E. Oliphant kirjoitti:
>  > Pauli Virtanen wrote:
>  [clip]
> >> I think this is a good idea: full-namespace docstring search à la Matlab
>  >> lookfor. I wrote a quick implementation for numpy here:
>  >> http://scipy.org/scipy/numpy/ticket/734
>  >>
>  >
>  > Cool.  I started scipy.misc.info a long time ago to try and do this.   I
>  > didn't advertise it well enough ;-)
>  >
>  > scipy.misc.info('eigvals')
>  >
>  > I'm happy to move this functionality into numpy (but it might be better
>  > in IPython).
>  Aha! I was wondering if something this simple was already present, but I
>  didn't notice that numpy.lib.utils.info also accepted strings as
>  parameters. Btw, the info function is already in numpy and the code is
>  duplicated in scipy.misc.
>  However, "info" is a bit different from "lookfor" in that it finds
>  documentation by the exact function name, not by looking into the
>  docstrings.
>  So, should info and lookfor be combined, or kept separate?
>  --
>  Pauli Virtanen
> _______________________________________________
>  Numpy-discussion mailing list
>  Numpy-discussion@scipy.org
>  http://projects.scipy.org/mailman/listinfo/numpy-discussion

Hoyt Koepke
UBC Department of Computer Science

More information about the Numpy-discussion mailing list