UKismetMathLibrary::Matrix_GetTransposeAdjoint
#include "Kismet/KismetMathLibrary.h"
Access: public
Specifiers: staticUFUNCTIONBlueprintPure
Description
Returns the transpose adjoint (cofactor matrix transposed) of the given matrix. Used in normal-vector transformations when non-uniform scaling is involved — multiply a normal by the transpose adjoint instead of the regular matrix to keep normals perpendicular to surfaces.
Caveats & Gotchas
- • The transpose adjoint is NOT the same as the inverse transpose. For an orthogonal matrix they coincide, but for matrices with shear or non-uniform scale they differ — make sure you want the adjoint, not UKismetMathLibrary::Matrix_GetInverse.
- • This operation involves computing all 9 cofactors of the 3×3 rotation sub-matrix and is comparatively expensive; avoid calling it every frame per mesh instance. Prefer pre-computing it when the matrix is known to be static.
Signature
static UE_INL_API FMatrix Matrix_GetTransposeAdjoint(const FMatrix& M)Parameters
| Name | Type | Description | Default |
|---|---|---|---|
| M | const FMatrix& | The input matrix. | — |
Return Type
FMatrix Example
Transform normals through a non-uniformly scaled matrix C++
FMatrix WorldMatrix = MyComponent->GetComponentTransform().ToMatrixWithScale();
FMatrix AdjointMat = UKismetMathLibrary::Matrix_GetTransposeAdjoint(WorldMatrix);
// Transform a local-space normal so it stays perpendicular after non-uniform scale
FVector LocalNormal = FVector(0.f, 0.f, 1.f);
FVector WorldNormal = AdjointMat.TransformVector(LocalNormal).GetSafeNormal();Version History
Introduced in: 4.0
| Version | Status | Notes |
|---|---|---|
| 5.6 | stable | — |
Feedback
Was this helpful?