Add step count display to JupyterViz#1775
Conversation
mesa/experimental/jupyter_viz.py
Outdated
| playing=playing.value, | ||
| on_playing=playing.set, | ||
| ) | ||
| widgets.IntText( |
There was a problem hiding this comment.
I tested, it looks like this
I think it shouldn't be boxed. Because a boxed UI element indicates it is user-editable.
There was a problem hiding this comment.
Instead of widgets.IntText, maybe try solara.Markdown instead.
Also, your code means that the widget is created from scratch every time on_value_play is called. A Solara widget should only be created once, and then is updated.
There was a problem hiding this comment.
@rht thanks for the feedback. I didn't see an obvious Solara output field so I followed the other widget examples; I had it set to disabled/read-only, but I agree a real output would be better and I can see now that markdown is feasible.
I've pushed an update that replaces widgets.IntText with solara.Markdown and tested locally, it is updating the number when I play through the simulation.
Codecov ReportPatch coverage has no change and project coverage change:
Additional details and impacted files@@ Coverage Diff @@
## main #1775 +/- ##
==========================================
- Coverage 80.29% 80.02% -0.28%
==========================================
Files 15 15
Lines 878 881 +3
Branches 187 188 +1
==========================================
Hits 705 705
- Misses 152 155 +3
Partials 21 21
☔ View full report in Codecov by Sentry. |
|
Sorry for the late reply. I tested. While the play button updates the step correctly, the step button doesn't. Try removing the |
|
I haven't tried, but what I meant was solara.Markdown(f"**step:** {viz.model.schedule.steps}, dependencies=[viz.model.schedule.steps]) |
|
I tried. solara.Markdown(f"**step:** {viz.model.schedule.steps}")worked like magic (no dependency needed). If you can confirm this. |
@rht yes, confirmed. It's great that it's so simple. I just pushed an update to my PR so that this the only line I'm proposing to change now (although it's really your solution, not mine!). |
|
Content looks good to me. The remaining "concern" is that the viz becomes slightly crowded. @tpike3 do you have concern about information density etc? If you are fine with it, then you can squash and merge this PR. |
Looks Good to Me --- thanks @rlskoeser! This is really good. |
3 participants
addresses #1773
I don't see any unit tests for the experimental JupyterViz code; I'd be game to add a test if/when there is existing testing infrastructure for this code, but not sure how to go about testing my contribution without examples for the other parts.
I'm not sure if there's any relevant documentation I need to update for this change. I could see if I can generate a new version of the screenshot for the Schelling segregration model displayed on the docs overview page that includes the step count, if that would be helpful/desirable.