12.21.2 Tensor Operation Type Verification

Tensor Operation Type Verification ensures correct algebraic manipulation by validating tensor operations within their defined mathematical frameworks.

Tensor Operation Type Verification is the check within the verification procedure that confirms the variance type of each index involved in a tensor operation, distinguishing contravariant indices from covariant indices, is compatible with the mathematical rules governing that operation before the operation is allowed to proceed.

Purpose of Type Verification

Preserving the Meaning of Variance

Contravariant and covariant indices transform differently under a change of basis, so type verification exists to prevent operations from combining components whose transformation behavior is incompatible, which would produce a result lacking a well-defined transformation law of its own.

Position Within the Broader Verification Procedure

Type verification is one of the specific checks carried out during the input verification stage of the overall tensor operation verification procedure, applied alongside other checks such as order and dimension agreement.

Checks Performed on Index Type

Matching Variance for Additive Operations

For operations such as addition and subtraction, type verification confirms that each pair of corresponding indices across the operands shares the same variance, since adding a contravariant component to a covariant component is not mathematically defined.

C = A + B

type (A^{i}) = type (B^{i})

Required Variance Pairing for Contraction

For contraction, type verification confirms that the index selected for summation appears once as an upper, contravariant index and once as a lower, covariant index, since the summation convention is defined only for such paired indices.

A_{i} B^{i} = \sum_{i = 1}^{n} A_{i} B^{i}

Unrestricted Variance for the Tensor Product

For the tensor product, type verification confirms only that the resulting object correctly concatenates the variance types of the two operands, since the tensor product itself imposes no requirement that the operand index types agree with one another.

(A \otimes B)^{i j} = A^{i} B^{j}

Outcomes of the Check

Successful Type Verification

When every index involved satisfies the variance requirement specific to the operation being performed, type verification passes and the procedure allows the remaining input checks, such as dimension agreement, to proceed.

Failure of Type Verification

When an index fails to satisfy the required variance pairing or matching, type verification fails and the procedure halts before computation, reporting the specific index and operation for which the variance requirement was not met.

Distinction from Related Checks

Difference from Order Verification

Type verification concerns the upper or lower placement of individual indices, while order verification concerns the total count of indices on a tensor; a tensor can satisfy order verification while still failing type verification if its indices carry the wrong variance for the operation.

Difference from Dimension Verification

Type verification concerns the qualitative role of an index as contravariant or covariant, while dimension verification concerns the quantitative size of the vector space associated with that index; both checks are applied independently to the same indices during input verification.

Relationship to Tensor Operation Notation

Type verification depends directly on the notational distinction between upper and lower indices established in tensor operation notation, since the placement of an index as a superscript or subscript is the explicit signal that the verification procedure reads to determine whether the variance requirements of a given operation are satisfied.