feat: Implement Slider class for JupyterViz

[rht]

This provides API parity with the previous Tornado viz. Supersedes #1945.
For the Schelling experimental example, from

model_params = {
    "density": {
        "type": "SliderFloat",
        "value": 0.8,
        "label": "Agent density",
        "min": 0.1,
        "max": 1.0,
        "step": 0.1,
    },
    "minority_pc": {
        "type": "SliderFloat",
        "value": 0.2,
        "label": "Fraction minority",
        "min": 0.0,
        "max": 1.0,
        "step": 0.05,
    },
    "homophily": {
        "type": "SliderInt",
        "value": 3,
        "label": "Homophily",
        "min": 0,
        "max": 8,
        "step": 1,
    },
    "width": 20,
    "height": 20,
}

to

model_params = {
    "density": Slider("Agent density", 0.8, 0.1, 1.0, 0.1),
    "minority_pc": Slider("Fraction minority", 0.2, 0.0, 1.0, 0.05),
    "homophily": Slider("Homophily", 3, 0, 8, 1),
    "width": 20,
    "height": 20,
}

I haven't added test because still waiting for agreement. But I have tested this manually on the Schelling experimental example.

[codecov]

Codecov Report

Attention: 9 lines in your changes are missing coverage. Please review.

Comparison is base (b60caa4) 79.27% compared to head (5f53869) 79.00%.
Report is 2 commits behind head on main.

Files	Patch %	Lines
mesa/experimental/jupyter_viz.py	33.33%	5 Missing and 1 partial ⚠️
mesa/experimental/UserParam.py	84.21%	2 Missing and 1 partial ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #1972      +/-   ##
==========================================
- Coverage   79.27%   79.00%   -0.28%     
==========================================
  Files          16       17       +1     
  Lines        1298     1324      +26     
  Branches      291      295       +4     
==========================================
+ Hits         1029     1046      +17     
- Misses        230      237       +7     
- Partials       39       41       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[quaquel]

I really like this API.

[rht]

Added test. This PR is ready for review.

[EwoutH]

Really like the API, this is good stuff.

Do we want to drop the explicit distinction between float and int? Is that implied with the values (mainly step) that you input now?

Will try to do a full review later today.

[quaquel]

Do we want to drop the explicit distinction between float and int? Is that implied with the values (mainly step) that you input now?

I am fine with dropping the distinction, but I can see arguments to have explicit int and float sliders. If desired, one possibility would be to just add a dtype kwarg to optionally control this behavior. If dtype is none, you derive the dtype from the arguments.

[rht]

Added an optional dtype argument.

[EwoutH]

@ankitk50 since you’re also working of visualization, what’s your take?

[ankitk50]

I would agree with being explicit with float/int data type. This would also later cater to the possibility of a custom slider with strings ['min', 'mid', 'max'].

[quaquel]

With the optional dtype argument, it should be quite easy to expand to categorical sliders with string labels later down the line. I can't fully judge the code because I am unfamiliar with solara. But from an API and extensibility point of view, this looks all fine to me.

[rlskoeser]

Really like how much more concise this is to configure.

[EwoutH]

Should it have an initial_value and a current_value (or just value)? So that you can reset it to it's original state, if wanted? (@Corvince would you think that is useful?)

[rht]

Should it have an initial_value and a current_value (or just value)? So that you can reset it to it's original state, if wanted?

That would need an extra button "reset to initial value" in order to be usable, and would be out of the scope of this PR. This PR wraps the same number of params in Solara's slider.

[quaquel]

I agree with @rth. Let's merge this.