---

name: makie-dynamic description: Makie animations, dashboards, and interactive visualizations using Observables, events, and UI widgets.

Makie Dynamic

Use Makie (not a specific backend) for all plotting. Assume all features are available.

Reactivity with Observables

• Use Observable values as the single source of truth for dynamic state.
• Prefer lift or @lift to derive dependent data reactively.
• After in-place mutation of an observable value (e.g. obs[] .= ...), call notify(obs).
• Use on(obs) for side effects and lift for pure transformations.

Animations and Recording

• For recording, use record(fig, "out.mp4", iterator) do frame ... end and update observables inside the loop.
• Avoid manual sleep loops or @async timers for animation.
• Use events(fig).tick for frame-aligned updates in interactive sessions.

UI Widgets

• Sliders: bind slider.value to observables; use @lift to connect to plot data.
• Buttons: use on(button.clicks) for actions (play/pause, reset, step).
• Toggles: gate updates by toggle.active[] to enable/disable behavior.
• Menus/Textboxes: use selection or value observables to drive state.

Events and Interaction

• Use events(fig) or events(ax) for mouse/keyboard/scroll events.
• If needed, return Consume(true) to stop lower-priority handlers.

Dashboards and Structure

• Keep dashboard state in a small set of observables (or a struct holding them).
• Split UI and plot construction into functions that accept GridPosition or Axis.
• For larger dashboards, consider using Makie.SpecApi to build declarative layouts.

Performance Notes

• Update plot data in-place when possible, and notify explicitly.
• Avoid rebuilding axes/plots each frame; update existing plot objects.
• For heavy pipelines, evaluate whether Makie’s compute pipeline tools are appropriate.