7-2. Key concepts

• Constraints apply to COG data.
• A constraint is a filter applied to the data — pixels that fall outside the constraint are masked out at render time.
• We already used the Constraint toggle on a layer to let users filter a layer by its own pixel values.
• Constraints can also come from secondary layers — for example land use, elevation, or "distance to" derived layers. Each constraint reads a separate COG and masks the primary layer where the constraint is not met.
• Secondary constraint layers must have the same CRS, resolution and origin as the primary data they constrain. This usually requires preparing compatible constraint layers in advance.
• Multiple constraints are applied together — a pixel is only drawn where every active constraint is satisfied.
• interactive: true exposes the control in the viewer so the user can change the filter; interactive: false fixes it.
• Each constraint is assigned a band index automatically, in the order the constraints are added.
• Constraint controls only appear in the viewer when the layer card control Constraint slider is enabled.

Constraint types

The GE defines three constraint types:

• Continuous — a variable (e.g. elevation) covering a full range between its min and max. Rendered as a slider.
• Categorical — data representing specific coded values (e.g. 10 = "Trees", 20 = "Grassland", …). Rendered as checkboxes.
• Combined — ranges within a continuous variable are grouped into named bands ("Low", "Medium", "High", "Very high") and rendered as checkboxes. Useful for things like altitudinal zones, aspect, uncertainty bands, or flood return periods.

See Constraints for the full reference.