Menu

Rainbow docs menu

Reference · Lemmas behind operations-research checks

rainbow.math.lemmas.dea-efficiency

Lemmas behind operations-research checks src/rainbow/math/lemmas/dea_efficiency.clj0 definitions, 2 theorems

General lemmas: dea efficiency. The mathematical facts the corresponding checks rest on, each checked by the kernel.

Depends onmath.sums.value-append math.nat.infinite math.rational.nnrat.identities math.nat.at-most-cancellation math.rational.nnrat.order-equivalence math.rational.nnrat.archimedean math.rational.nnrat.add-monotone math.rational.nnrat.add-cancellation math.rational.nnrat.additive-reassociation math.rational.signed.order-transitivity math.rational.signed.convergence math.rational.signed.order-respects math.rational.signed.limit-uniqueness math.rational.signed.strict-order math.rational.signed.multiply math.rational.nnrat.multiply-commute math.rational.nnrat.product-reassociation math.rational.nnrat.distribute-right math.rational.nnrat.ring math.rational.nnrat.multiply-monotone math.rational.nnrat.order-difference math.rational.signed.multiply-order math.rational.signed.equivalence-laws math.rational.nnrat.minmax math.rational.nnrat.farkas-elim-list math.rational.signed.no-zero-divisors math.rational.signed.decide-order math.rational.signed.decide-reflexive math.logic.boolean-and math.rational.signed.decide-equivalence math.rational.signed.decide-completeness math.rational.signed.unit-laws math.nat.strict-below-split math.nat.strict-below-more math.checks math.nat.at-most-add math.rational.nnrat.harmonic-convergence math.nat.divmod math.rational.signed.minmax math.rational.signed.minmax-algebra math.rational.signed.clamp math.rational.signed.deviation-rows math.rational.signed.absolute math.rational.signed.absolute-algebra math.rational.signed.distance math.nat.linear-order math.rational.signed.add-associative math.nat.semiring math.nat.multiply-zero-split math.rational.nnrat.multiply-zero-split math.nat.monus math.rational.nnrat.monus math.rational.nnrat.add-zero-split math.rational.nnrat.add-cancel-zero math.finite.rational-argmax math.rational.nnrat.order-total math.rational.nnrat.multiply-difference-zero math.linear.vector.norm-nonnegative math.linear.vector.norm-zero-reflection algorithms.proofs.counterfactual-distance-separation algorithms.proofs.counterfactual-distance-laws math.finite.matrix math.linear.matrix.equivalence math.linear.matrix.transpose math.linear.matrix.transpose-equivalence math.linear.matrix.multiply math.linear.matrix.multiply-shape math.linear.matrix.composable math.linear.matrix.composable-multiply math.linear.vector.additive-units math.linear.vector.shape-cancellation math.linear.vector.shape-map math.linear.vector.shape-add math.linear.vector.dot-additive math.linear.vector.dot-bilinear math.linear.vector.dot-negate math.linear.vector.dot-subtract math.linear.vector.dot-scale math.linear.matrix.action-scale math.rational.signed.multiply-nonzero math.rational.signed.order-monotone math.rational.signed.list-sum math.rational.signed.multiply-monotone math.rational.signed.map-sum-algebra math.rational.signed.negate-additive math.rational.signed.lp-objective-negate math.nat.pair-index-inverse math.sums.flow-cut math.rational.signed.multiply-associative math.rational.signed.multiply-ring

signed/bilinear-swap

TheoremFor all inputsNo axiomsSource

signed/bilinear-swap :
  (a b : ℕ → ℚ) (M : ℕ → ℕ → ℚ) (p q : ℕ)
  → signed/expr-sum-value (fun l1 => a l1 · signed/expr-sum-value (fun ib => b ib · M l1 ib) q) p = signed/expr-sum-value (fun r1 => b r1 · signed/expr-sum-value (fun ia => a ia · M ia r1) p) q

= = signed/equivalent: equality of rational values· = signed/multiplyℕ = data/Natℚ = signed/SignedRational

signed/dea-weak-duality

TheoremFor all inputsNo axiomsSource

signed/dea-weak-duality :
  (X Y : ℕ → ℕ → ℚ) (u v lam : ℕ → ℚ) (th : ℚ) (m s n : ℕ)
  (hu : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx s), 0 ≤ u idx)
  (hv : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx m), 0 ≤ v idx)
  (hnorm : signed/expr-sum-value (fun wk => v wk · X wk 0) m = 1)
  (hmult : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx n), signed/expr-sum-value (fun wk => u wk · Y wk idx) s ≤ signed/expr-sum-value (fun wk => v wk · X wk idx) m)
  (hlam : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx n), 0 ≤ lam idx)
  (hin : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx m), signed/expr-sum-value (fun lk => lam lk · X idx lk) n ≤ th · X idx 0)
  (hout : ∀ (idx : ℕ), ∀ (bound : nat/strict-below idx s), Y idx 0 ≤ signed/expr-sum-value (fun lk => lam lk · Y idx lk) n)
  → signed/expr-sum-value (fun wk => u wk · Y wk 0) s ≤ th

0 = signed/zero1 = signed/one= = signed/equivalent: equality of rational values· = signed/multiplyℕ = data/Natℚ = signed/SignedRational≤ = signed/at-most