In this lesson, you will learn how to create a responsive radial menu that maps circular mouse movement to precise selection states using custom input routing.
The Core Concept: Angular Input Mapping
Unlike standard list menus, a radial menu relies on calculating the angle between the center of the widget and the user's cursor position. Instead of relying on OnClicked events, we handle OnMouseMove within the widget to dynamically update the active slice based on the current degree range.
To achieve this, we translate the mouse coordinate relative to the widget's center using (Cursor - Center). We then use the Atan2(y, x) function to return the angle in radians, which we convert to degrees.
// Logic inside our Custom Radial Widget
FVector2D Center = GetTickSpaceGeometry().GetAbsoluteSize() / 2.0f;
FVector2D Delta = MousePosition - Center;
float Angle = FMath::RadiansToDegrees(FMath::Atan2(Delta.Y, Delta.X));
// Normalize to 0-360 range
float NormalizedAngle = FMath::Fmod(Angle + 360.0f, 360.0f);
State-Driven Selection Animations
Fluidity comes from decoupling the "Active Index" from the visual representation. When the calculated NormalizedAngle shifts into a new slice, we trigger a state change. Rather than instantly snapping, we use a Widget Blueprint Animation or a Lerp function within the Tick (or an Invalidation Panel) to transition the slice's scale and color.
By implementing custom input routing, we ensure that the radial menu consumes the mouse movement while ignoring standard focus events. We achieve this by overriding OnMouseMove and returning FReply::Handled(). If the distance from the center is below a certain "deadzone" threshold, we maintain the "Center" state, effectively creating a "Cancel" or "No-Selection" behavior that feels intuitive to the player.
Combining these mathematical mappings with state-driven transitions ensures that as the mouse circles the screen, the UI feels like a physical, reactive extension of the player's intent rather than a static list of buttons.
Key Takeaway: Radial menus succeed by mapping mouse vectors to angular slices and driving visual states through smooth interpolation rather than binary selection.