IndisputableMonolith.Information.Compression
The Information.Compression module links data compression to J-cost minimization in Recognition Science. It defines the base-2 logarithm and establishes the source coding theorem that bounds achievable rates by Shannon entropy. Supporting definitions distinguish message J-cost, lossless compression, and lossy compression. Information theorists working inside the RS framework would cite these results for entropy-derived compression limits.
claimDefines the base-2 logarithm $log_2 x$ and states the source coding theorem: for a discrete source with Shannon entropy $H = -∑ p_i log_2 p_i$, the minimal average code length per symbol satisfies $L ≥ H$, with equality in the asymptotic limit for lossless schemes.
background
The module belongs to the Information domain and imports the RS time quantum τ₀ = 1 tick from Constants together with J-cost primitives from the Cost module. It extends the Shannon entropy construction (INFO-001) that derives H = -∑ p_i log p_i directly from J-cost structure. Core objects include sourceEntropy for an arbitrary probability distribution, fairCoinEntropy for the uniform binary case, and biasedCoinEntropy for nonuniform bits.
proof idea
This is a definition module containing supporting lemmas. Logarithmic identities (log2_two, log2_half) are established first from Mathlib properties. The source_coding_theorem is obtained by applying the entropy definition to average code lengths. The key equivalence compression_is_jcost_minimization follows by expressing code length as a J-cost sum, after which messageJCost, losslessCompression, and lossyCompression are introduced as definitions.
why it matters in Recognition Science
The module supplies the compression layer that converts Shannon entropy into concrete rate bounds inside Recognition Science. It aligns with J-uniqueness (T5) by casting compression as J-cost minimization and prepares the ground for efficiency measures that later connect to the phi-ladder. No downstream theorems are yet listed, so the module functions as a foundational block for the broader information framework.
scope and limits
- Does not treat noisy channels or error-correcting codes.
- Does not supply explicit compression algorithms.
- Does not address continuous sources or differential entropy.
- Does not relate compression rates to the phi-ladder or spatial dimension D=3.
depends on (3)
declarations in this module (23)
-
def
log2 -
lemma
log2_two -
lemma
log2_half -
theorem
source_coding_theorem -
def
sourceEntropy -
def
fairCoinEntropy -
theorem
fair_coin_one_bit -
def
biasedCoinEntropy -
theorem
compression_is_jcost_minimization -
def
messageJCost -
def
losslessCompression -
def
lossyCompression -
def
rateDistortionTheory -
def
kolmogorovComplexity -
theorem
most_strings_incompressible -
def
huffmanCoding -
def
arithmeticCoding -
def
lempelZiv -
def
englishLetterEntropy -
def
englishEntropyRate -
theorem
english_is_redundant -
def
summary -
structure
CompressionFalsifier