pith. machine review for the scientific record. sign in
def definition def or abbrev high

lossyCompression

show as:
view Lean formalization →

The lossyCompression definition supplies a string label for approximate reconstruction that accepts distortion, allowing rates below source entropy. Information theorists cite it when extending Shannon limits to rate-distortion trade-offs inside the Recognition Science J-cost framework. The realization is a direct string assignment with no computation or lemmas.

claimLossy compression is defined as approximate reconstruction that accepts distortion $D$, permitting a rate $R(D)$ below the entropy $H(X)$ by discarding high $J$-cost information.

background

The Information.Compression module derives compression limits from J-cost. Shannon's source coding theorem requires average code length at least $H(X) = -∑ p(x) log₂ p(x)$. In Recognition Science, every recognition event carries nonnegative J-cost, and compression lowers total J-cost by increasing organization.

proof idea

Direct string definition assigning the literal value. No lemmas from upstream cost definitions or forcing structures are invoked; the body is a one-line string literal.

why it matters in Recognition Science

This definition completes the lossy case inside INFO-003 on data compression limits from J-cost. It supports the module claim that compression equals J-cost minimization and prepares extensions toward rate-distortion theory while remaining consistent with the Recognition Composition Law.

scope and limits

formal statement (Lean)

 137def lossyCompression : String :=

proof body

Definition body.

 138  "Approximate reconstruction, accepts distortion"
 139
 140/-- Rate-distortion theory:
 141
 142    R(D) = minimum rate for distortion ≤ D
 143
 144    Trade-off between compression and quality.
 145
 146    In RS: Which ledger information to discard
 147    based on J-cost importance. -/

depends on (4)

Lean names referenced from this declaration's body.