A mathematical transformation that converts a quantity, vector, operator, or expression from one representational basis or coordinate system to another so that the same underlying object can be related across formulations.

Principle

Use an invertible mapping (often linear) between isomorphic vector spaces or algebraic structures to express objects in the basis best suited for a task while preserving intrinsic relations and invariant properties.

Demonstration

Changing a vector's coordinates from the standard basis to an orthonormal eigenbasis via matrix P so that x' = P^{-1} x; or applying the Fourier transform to move from time domain to frequency domain to simplify convolution into multiplication.

Misapplication

Applying a non‑invertible or ill‑defined transform between non‑isomorphic spaces (for instance treating truncated coefficients as a lossless change of basis) which destroys information or invalidates invariants.

Consequence

A valid change of representation reveals structure (diagonalizes operators, separates scales), enables more efficient computation or approximation, and allows comparison of formulations that are equivalent up to the transform.

Reversal

Returning to the original representation by applying the inverse transform; conceptually, the reversal is a representation-preserving identity when the mapping is invertible.

Boundary

Requires a well‑defined mapping between compatible spaces and attention to domains, boundary conditions, and completeness; it excludes arbitrary re‑ encodings that change the semantics or topology of the object.

Semantic Tension

Tension exists between change of representation (a formal coordinate or basis transform preserving identity) and change of model (a different mathematical description that may approximate or alter the object); the former preserves equivalence, the latter may not.

Synthesis

A change of representation is an invertible mapping that re-expresses the same mathematical object in a different basis or coordinates to expose structure, simplify operations, or relate distinct formulations while preserving core invariants.