Module dimensional 

Dimensional Configuration - Variable Precision Hyperdimensional Substrate

This module implements configurable dimensional encoding where:

• Number of dimensions is tunable (sparse vs dense tradeoff)
• Trit depth per dimension is variable (precision vs storage tradeoff)
• Differential encoding reduces actual storage to deltas from codebook

Mathematical Foundation

Each dimension d has a configurable trit depth T_d:

• 1 trit = 3 states {-1, 0, +1} = 1.585 bits
• 2 trits = 9 states = 3.17 bits
• 3 trits = 27 states = 4.75 bits
• N trits = 3^N states = N × 1.585 bits

Total information capacity: Σ(T_d × 1.585) bits across all dimensions

Algebraic Primitives

These operations form the low-level compute substrate:

• Bundle (⊕): Superposition - trit-wise majority voting
• Bind (⊙): Composition - trit-wise multiplication
• Permute (ρ): Sequence encoding - cyclic index shift
• Cosine: Similarity - normalized dot product
• Factorize: Decomposition - resonator-based unbinding

All operations preserve balanced ternary properties and are algebraically closed.

Structs

DifferentialEncoder

Differential encoder using a codebook basis

DifferentialEncoding

Differential encoding result

DimensionalConfig

Configuration for dimensional encoding

HyperVec

A hyperdimensional vector with configurable dimensional depth

Tryte

Tryte: A group of trits (configurable size) Default is 6 trits = 729 states ≈ 9.51 bits

Enums

Trit

Single trit: balanced ternary digit {-1, 0, +1}

TritDepthConfig

Configuration for trit depth across dimensions