13.21.4 Tensor Contraction Trace Notation

Tensor Contraction Trace Notation is a method to simplify tensor expressions by summing over repeated indices, commonly used in physics and mathematics.

Tensor Contraction Trace Notation is the compact written form used specifically to denote the full self-contraction of a rank-2 mixed tensor over its single upper-lower index pair, replacing the general repeated index notation for this particular case with a dedicated symbol that emphasizes the scalar, basis-independent nature of the resulting quantity.

Definition

For a rank-2 mixed tensor $T_{a}^{b}$ , trace notation denotes its full contraction as:

tr (T) = T_{a}^{a}

replacing the repeated-index expression on the right with the dedicated symbol $tr (T)$ on the left, both denoting the identical scalar value.

Why a Dedicated Symbol Is Used

Emphasizing Scalar Status

Because a full self-contraction of a rank-2 mixed tensor always yields a scalar, trace notation immediately signals to the reader that the result is a basis-independent number, without requiring inspection of the index pattern to confirm that no free indices remain.

Convenience for Repeated Use

The trace operation recurs frequently across tensor algebra, appearing in contexts ranging from matrix theory to curvature computations, and the dedicated notation avoids repeatedly rewriting the full repeated-index form each time the operation is invoked.

Properties Expressed Through the Notation

Linearity

Trace notation makes the linearity of the trace operation immediately transparent:

tr (α S + β T) = α tr (S) + β tr (T)

for scalars $α$ and $β$ , following directly from the linearity of the underlying summation.

Cyclic Property for Products

For a product of two rank-2 mixed tensors, trace notation expresses the cyclic invariance property compactly:

tr (A B) = tr (B A)

even though the individual repeated-index expressions for $A B$ and $B A$ appear, before simplification, to differ in the order of their factors.

Diagram

Position Within Notation

Trace notation is a specialized abbreviation within the broader tensor contraction notation, reserved specifically for the case of a full rank-2 self-contraction, and it corresponds directly to the closed loop of length one in the diagrammatic representation, offering a written shorthand for exactly the same self-loop structure that the diagram depicts visually.