Refactor and improve patching of the plotting methods docstring and signature #1520
Conversation
hoxbro
left a comment
hoxbro
left a comment
There was a problem hiding this comment.
I'm assuming most of the changes are "just" moved code, and no significant logic has been changed in them.
| from .util import _PatchHvplotDocstrings, _in_ipython | ||
| from .utilities import help, hvplot_extension, output, save, show # noqa |
There was a problem hiding this comment.
It's a little weird that there are two files with util/utilities, but this was not added to this PR.
There was a problem hiding this comment.
That's the sort of thing we should clean up for hvPlot 1.0.
hvplot/util.py
Outdated
| return geoviews | ||
|
|
||
|
|
||
| def _is_installed(package: str) -> bool: |
There was a problem hiding this comment.
| def _is_installed(package: str) -> bool: | |
| def _is_installed(module: str) -> bool: |
I see a module as something you import and a package as something you install. In most cases, this is, of course, the same.
There was a problem hiding this comment.
Yes thanks good catch, I didn't know find_spec was looking at modules. I removed _is_installed entirely as I think it was a bit misleading and isn't used much anyway.
There was a problem hiding this comment.
Maybe cache this result?
I thought about it before, and then thought people do special things in their environment while working directly from a notebook (install packages, autoreloading modules, etc.). The function that patches plotting signatures and docstrings doesn't only run on import but also when switching HoloViews backend. So I'd rather not cache for now, and keep that for future improvements to improve the import speed of hvPlot, which I'm sure we'll look at at some point.
HVPLOT_PATCH_PLOT_DOCSTRING_SIGNATUREis non falsy__init__.pyto theutil.pymodule