hgrecco / pint / 1868

92%

master: 91%

push

travis-ci

Merge #1133

1133: Change HTML repr away from LaTeX to a Sparse/Dask-like repr r=hgrecco a=jthielen

While I've never been a big fan of Pint's LaTeX repr in JupyterLab/Jupyter Notebook due to working with large arrays that would crash the renderer, I've recently been working a lot with various wrapping combinations of xarray, Pint, Dask, and Sparse, in which Pint's LaTeX repr performs particularly poorly. Even if Pint's repr were elided (#765), it would still poorly nest with the other HTML reprs.

This motivated me to implement an alternative HTML repr for Pint using a simple HTML table similar to what both Dask and Sparse use. As shown in [this demonstration notebook](https://nbviewer.jupyter.org/urls/dl.dropbox.com/s/jlgqm6s92kzvagi/pint_html_repr_demo.ipynb), this allows much better nesting of HTML reprs. A quick example image is below:

![](https://www.meteor.iastate.edu/~jthielen/Screenshot%20from%202020-07-14%2017-07-51.png)

This doesn't quite take care of [xarray's unit repr discussions](https://github.com/pydata/xarray/issues/2773) (since this only shows up when expanding the data in a dataset), but I still think it is a worthwhile improvement for the expanded view. @keewis do you have any thoughts on this particular interaction?

Marking as a WIP/draft for now to see if this is a welcome format for the HTML repr first before I dig too much into optimizing the implementation and writing tests for it.

- [x] Closes #654 (even though it was already closed in favor of #765, I think the are related but separate issues)
- [x] Executed ``black -t py36 . && isort -rc . && flake8`` with no errors
- [x] The change is fully covered by automated unit tests
- [x] Documented in docs/ as appropriate
- [x] Added an entry to the CHANGES file

Tagging @nbren12, @hgrecco, @jondoesntgit, @natezb, @ipcoder, @keewis, and @TomNicholas for input based on past issues (#654, #765, and https://github.com/xarray-contrib/pint-xarray/issues/6).


Co-aut... (continued)

25 of 25 new or added lines in 2 files covered. (100.0%)

3649 of 4072 relevant lines covered (89.61%)

11.79 hits per line