12.22 Tensor Operation Boundary

Tensor Operation Boundary is the set of mathematical limits that determine which combinations of tensors, indices, and dimensions a given operation may legitimately act upon, marking the point beyond which an operation ceases to be defined or produces a result inconsistent with the algebraic structure of tensors.

Nature of the Boundary

A Limit on Applicability

The boundary of a tensor operation is not a numerical bound on component values but a structural limit on the combinations of order, index variance, and dimension for which the operation retains a well-defined meaning.

Distinguishing Defined from Undefined Cases

Every tensor operation partitions the space of possible operand combinations into those for which the operation is defined and those for which it is not, and the operation boundary is precisely the dividing line between these two regions.

Boundaries Associated with Specific Operations

Boundary of Addition

Addition is bounded to operands sharing identical order, matching index variance in every corresponding position, and equal dimension along every corresponding index, so that combinations violating any of these conditions fall outside the boundary of the operation.

Boundary of Contraction

Contraction is bounded to a chosen pair of indices consisting of exactly one contravariant and one covariant index sharing the same dimension, so that attempting to contract two indices of the same variance, or of differing dimension, lies outside the operation's boundary.

Boundary of the Tensor Product

The tensor product carries the widest boundary among common operations, remaining defined for operands of any order and any combination of variance, since it does not require matching between the operands but only well-defined vector spaces underlying each factor.

Function of the Boundary in the Verification Procedure

Reference Standard for Verification Checks

The boundary of an operation supplies the reference standard against which the input verification stage tests supplied operands, since each specific check performed during input verification, such as order or dimension agreement, corresponds to one condition defining the operation's boundary.

Consequence of Operating Outside the Boundary

An attempt to apply an operation to operands lying outside its boundary is rejected during input verification rather than being carried out and later judged incorrect, since the boundary marks the operation as undefined rather than merely producing an unusual result.

Boundary as a Fixed Property of the Operation

Independence from Specific Component Values

The boundary of an operation is determined entirely by structural properties, order, variance, and dimension, and does not depend on the particular numerical values held by the components of the tensors involved.

Stability Across Instances of the Same Operation

Because the boundary depends only on structural properties, it remains the same for every instance of a given operation, allowing the same set of verification checks to be applied uniformly whenever that operation is invoked.

Relationship to Tensor Operation Notation

The conditions defining an operation's boundary are expressed using the same indicial notation used to describe tensors generally, so that the placement and count of upper and lower indices in an expression directly reveal whether a proposed application of an operation falls inside or outside its defined boundary.