15.4 Tensor Symmetric Component Constraint Pattern

Tensor Symmetric Component Constraint Pattern is the recurring structural pattern by which symmetry constraints organize themselves across rank and dimension, revealing that the equations forcing components to agree are not scattered arbitrarily but follow a predictable, recursive arrangement tied directly to how index multisets are built up one index at a time. Where an individual constraint signal states a single equality between two components, the constraint pattern describes the overall shape formed by all such equalities taken together.

The Recursive Building Pattern

Adding One Index at a Time

The pattern underlying symmetric constraints becomes visible by building a rank-( r ) symmetric tensor from a rank-( (r-1) ) one: each independent component of the rank-( r ) tensor is obtained by appending one additional index to an existing independent component of lower rank, provided the appended index is no smaller than the last index already present. This ordering rule is what prevents the same equivalence class from being counted more than once.

Recursive Count Relation

This recursive construction produces a relation between successive symmetric rank areas:

which is the same identity that generates Pascal's triangle, showing that the growth of independent symmetric components from one rank to the next follows exactly this additive branching pattern rather than an arbitrary count.

Nondecreasing Index Sequences

The Canonical Representative Pattern

Every equivalence class of indices produced by symmetry constraints has exactly one representative written in nondecreasing order:

and the constraint pattern can be read as a rule that collapses every ordered tuple onto this single sorted representative. This nondecreasing sequence is the pattern that all constraint signals, taken together, ultimately reduce to.

Diagram of the Sorted Pattern

Pattern Across a Fixed Multiset

Multiplicity Within the Pattern

When the index multiset contains repeated values, the constraint pattern still produces exactly one independent component, but the number of ordered tuples collapsing onto it varies according to a multinomial count. For a multiset with multiplicities ( \alpha_1, \ldots, \alpha_n ) summing to ( r ), the number of ordered arrangements folded together by the pattern is:

so multisets with more repeated values are collapsed by a smaller pattern block, while multisets of all-distinct indices are collapsed by the largest possible block of size ( r! ).

Layered Pattern by Repetition Type

Grouping index multisets by their repetition structure produces layers within the constraint pattern: multisets with no repeated index sit in the layer with the largest collapsing block, multisets with exactly one pair repeated sit in an intermediate layer, and the fully repeated multiset, where every index equals the same value, sits in the layer with a trivial collapsing block of size one, since no permutation changes such a tuple at all.

Consistency With the Overall Structure

Pattern Matches the Rank Area Total

Summing the collapsing block sizes across every distinct multiset reproduces the total raw component count ( n^r ), while the number of distinct multisets themselves reproduces the symmetric rank area dimension, confirming that the constraint pattern precisely partitions the full tensor space into the equivalence classes counted by the rank area formula.

Pattern as a Generating Tool

Because the pattern is recursive and follows the nondecreasing-sequence rule, it can be used directly to enumerate a complete, non-redundant basis for a symmetric tensor: listing every nondecreasing sequence of length ( r ) drawn from ( n ) values produces the full and exact set of independent components predicted by the pattern, with no component missing and none duplicated.