✦ For everyone, free.

Practical knowledge for real and everyday life

Home

10.5 Tensor Forward Basis Change Rule

The Tensor Forward Basis Change Rule describes how tensor components transform when switching from one basis to another in a vector space.

Tensor Forward Basis Change Rule is the formal statement fixing, once and for all, which matrix factor applies to each tensor index when moving from a designated source basis to a designated target basis, serving as the canonical reference rule from which every other direction, area, and special case of basis transformation is derived.


Statement of the Rule

The Basis Vectors Under the Forward Rule

The forward rule begins by fixing how the target basis vectors are expressed in terms of the source basis vectors, using the matrix (A) as the defining forward relation.

ei = Aij ej

Contravariant Components Under the Forward Rule

The forward rule then fixes, as a direct logical consequence of preserving the underlying tensor, that contravariant components transform using (A^{-1}) in this same forward direction.

vi = (A1) j i vj

Covariant Components Under the Forward Rule

Correspondingly, the forward rule fixes covariant components to transform using (A) directly, matching the basis vectors' own transformation.

ωi = Aij ωj

Role of the Rule as a Reference Convention

Establishing the Baseline for Every Other Case

Every alternative arrangement, such as reversing which basis is called the source, adopting the opposite column convention for the matrix, or extending to Jacobian-based coordinate transformations, is defined and understood relative to this single forward rule serving as the fixed reference point.

Why the Rule Is Called "Forward"

The rule is labeled forward specifically because it proceeds in the direction from a basis already known, the source, toward a basis being sought, the target, mirroring the natural order in which a practitioner typically approaches a basis change problem: starting from known data and computing toward a desired result.


Generalization of the Forward Rule

Extension to Mixed Tensors of Any Rank

The forward rule extends uniformly to a tensor with any number of contravariant and covariant indices, applying one factor of (A^{-1}) to each contravariant index and one factor of (A) to each covariant index, following the same pattern established for vectors and covectors individually.

Tlk = (A1) i k Alj Tji

Extension to the Coordinate Setting

Replacing the constant matrix (A) with a position-dependent Jacobian extends the forward rule directly to curvilinear coordinate transformations, without altering the fundamental combinatorial pattern of one matrix or inverse factor per index.


Verifying the Forward Rule Preserves the Tensor

The Substitution Check

Substituting the forward rule for both the basis vectors and the components into the invariance identity confirms that the abstract tensor remains unchanged, which is the ultimate justification for why the rule takes precisely this form and no other.

vi ei = vi ei

Visual Illustration

Source basis (known) forward rule Target basis (sought) Basis: apply A Contravariant components: apply A inverse Covariant components: apply A

Why a Single Canonical Rule Matters

Fixing one forward rule as the canonical reference is what allows every related notion, reverse transformations, alternative matrix conventions, and coordinate generalizations, to be defined precisely as variations on a single, unambiguous baseline. Without this canonical statement, discussions of basis change would risk sliding between inconsistent conventions without a clear common reference point to return to.

Content in this section