Consistent visualization theme #974
Replies: 4 comments 1 reply
|
The dpi in arviz can be ajjusted globaly using rcparams. Additionally, I have been thinking defaulting to a smaller value. Not sure what do you mean by "how do we use it in non-arviz visualizations?" The themes are global, affecting matplotlib, plotly and bokeh. I suggest to use an arviz-theme, modifications to existing themes or new themes are welcome. |
|
Is this the only page with arviz available styles? https://python.arviz.org/en/stable/examples/styles.html From the list above, I prefer arviz-docgrid. But I'd check arviz-variat as well. |
|
I think we are missing a page for the new styles. These are the styles https://arviz-devs.github.io/EABM/Chapters/Elements_of_visualization.html#style-sheets When I said modifications to existing themes are welcome, I mean modifications to the styles in arviz. But it is also possible to apply customizations for Bambi docs. For instance, adding a grid. |
Uh oh!
There was an error while loading. Please reload this page.
In our documentation we have visualizations created with several different themes:
bambi.interpretare built on top ofseaborn.objects.I would like us to have visualizations in our documentation follow a consistent theme.
Should that be...
Finally, even though I like seaborn.object's API, I don't think we should give users the impression that they need to use seaborn to customize a visualization created with
bambi.interpret.plot_*. It's fine if we do that in notebooks, but there should be at least one example showing how to perform customizations with plain matplotlib.All reactions