P. 392 of Theorem List - Metamath Proof Explorer

Theorem List for Metamath Proof Explorer - 39101-39200   *Has distinct variable group(s)

Type	Label	Description
Statement
Theorem	jm2.27b 39101	Lemma for jm2.27 39103. Expand existential quantifiers for reverse direction. (Contributed by Stefan O'Rear, 4-Oct-2014.)
⊢ (𝜑 → 𝐴 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝐵 ∈ ℕ)    &   ⊢ (𝜑 → 𝐶 ∈ ℕ)    &   ⊢ (𝜑 → 𝐷 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐸 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐹 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐺 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐻 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐼 ∈ ℕ0)    &   ⊢ (𝜑 → 𝐽 ∈ ℕ0)    &   ⊢ (𝜑 → ((𝐷↑2) − (((𝐴↑2) − 1) · (𝐶↑2))) = 1)    &   ⊢ (𝜑 → ((𝐹↑2) − (((𝐴↑2) − 1) · (𝐸↑2))) = 1)    &   ⊢ (𝜑 → 𝐺 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → ((𝐼↑2) − (((𝐺↑2) − 1) · (𝐻↑2))) = 1)    &   ⊢ (𝜑 → 𝐸 = ((𝐽 + 1) · (2 · (𝐶↑2))))    &   ⊢ (𝜑 → 𝐹 ∥ (𝐺 − 𝐴))    &   ⊢ (𝜑 → (2 · 𝐶) ∥ (𝐺 − 1))    &   ⊢ (𝜑 → 𝐹 ∥ (𝐻 − 𝐶))    &   ⊢ (𝜑 → (2 · 𝐶) ∥ (𝐻 − 𝐵))    &   ⊢ (𝜑 → 𝐵 ≤ 𝐶)    ⇒   ⊢ (𝜑 → 𝐶 = (𝐴 Yrm 𝐵))
Theorem	jm2.27c 39102	Lemma for jm2.27 39103. Forward direction with substitutions. (Contributed by Stefan O'Rear, 4-Oct-2014.)
⊢ (𝜑 → 𝐴 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝐵 ∈ ℕ)    &   ⊢ (𝜑 → 𝐶 ∈ ℕ)    &   ⊢ (𝜑 → 𝐶 = (𝐴 Yrm 𝐵))    &   ⊢ 𝐷 = (𝐴 Xrm 𝐵)    &   ⊢ 𝑄 = (𝐵 · (𝐴 Yrm 𝐵))    &   ⊢ 𝐸 = (𝐴 Yrm (2 · 𝑄))    &   ⊢ 𝐹 = (𝐴 Xrm (2 · 𝑄))    &   ⊢ 𝐺 = (𝐴 + ((𝐹↑2) · ((𝐹↑2) − 𝐴)))    &   ⊢ 𝐻 = (𝐺 Yrm 𝐵)    &   ⊢ 𝐼 = (𝐺 Xrm 𝐵)    &   ⊢ 𝐽 = ((𝐸 / (2 · (𝐶↑2))) − 1)    ⇒   ⊢ (𝜑 → (((𝐷 ∈ ℕ0 ∧ 𝐸 ∈ ℕ0 ∧ 𝐹 ∈ ℕ0) ∧ (𝐺 ∈ ℕ0 ∧ 𝐻 ∈ ℕ0 ∧ 𝐼 ∈ ℕ0)) ∧ (𝐽 ∈ ℕ0 ∧ (((((𝐷↑2) − (((𝐴↑2) − 1) · (𝐶↑2))) = 1 ∧ ((𝐹↑2) − (((𝐴↑2) − 1) · (𝐸↑2))) = 1 ∧ 𝐺 ∈ (ℤ≥‘2)) ∧ (((𝐼↑2) − (((𝐺↑2) − 1) · (𝐻↑2))) = 1 ∧ 𝐸 = ((𝐽 + 1) · (2 · (𝐶↑2))) ∧ 𝐹 ∥ (𝐺 − 𝐴))) ∧ (((2 · 𝐶) ∥ (𝐺 − 1) ∧ 𝐹 ∥ (𝐻 − 𝐶)) ∧ ((2 · 𝐶) ∥ (𝐻 − 𝐵) ∧ 𝐵 ≤ 𝐶))))))
Theorem	jm2.27 39103*	Lemma 2.27 of [JonesMatijasevic] p. 697; rmY is a diophantine relation. 0 was excluded from the range of B and the lower limit of G was imposed because the source proof does not seem to work otherwise; quite possible I'm just missing something. The source proof uses both i and I; i has been changed to j to avoid collision. This theorem is basically nothing but substitution instances, all the work is done in jm2.27a 39100 and jm2.27c 39102. Once Diophantine relations have been defined, the content of the theorem is "rmY is Diophantine" (Contributed by Stefan O'Rear, 4-Oct-2014.)
⊢ ((𝐴 ∈ (ℤ≥‘2) ∧ 𝐵 ∈ ℕ ∧ 𝐶 ∈ ℕ) → (𝐶 = (𝐴 Yrm 𝐵) ↔ ∃𝑑 ∈ ℕ0 ∃𝑒 ∈ ℕ0 ∃𝑓 ∈ ℕ0 ∃𝑔 ∈ ℕ0 ∃ℎ ∈ ℕ0 ∃𝑖 ∈ ℕ0 ∃𝑗 ∈ ℕ0 (((((𝑑↑2) − (((𝐴↑2) − 1) · (𝐶↑2))) = 1 ∧ ((𝑓↑2) − (((𝐴↑2) − 1) · (𝑒↑2))) = 1 ∧ 𝑔 ∈ (ℤ≥‘2)) ∧ (((𝑖↑2) − (((𝑔↑2) − 1) · (ℎ↑2))) = 1 ∧ 𝑒 = ((𝑗 + 1) · (2 · (𝐶↑2))) ∧ 𝑓 ∥ (𝑔 − 𝐴))) ∧ (((2 · 𝐶) ∥ (𝑔 − 1) ∧ 𝑓 ∥ (ℎ − 𝐶)) ∧ ((2 · 𝐶) ∥ (ℎ − 𝐵) ∧ 𝐵 ≤ 𝐶)))))
Theorem	jm2.27dlem1 39104*	Lemma for rmydioph 39109. Substitution of a tuple restriction into a projection that doesn't care. (Contributed by Stefan O'Rear, 11-Oct-2014.)
⊢ 𝐴 ∈ (1...𝐵)    ⇒   ⊢ (𝑎 = (𝑏 ↾ (1...𝐵)) → (𝑎‘𝐴) = (𝑏‘𝐴))
Theorem	jm2.27dlem2 39105	Lemma for rmydioph 39109. This theorem is used along with the next three to efficiently infer steps like 7 ∈ (1...;10). (Contributed by Stefan O'Rear, 11-Oct-2014.)
⊢ 𝐴 ∈ (1...𝐵)    &   ⊢ 𝐶 = (𝐵 + 1)    &   ⊢ 𝐵 ∈ ℕ    ⇒   ⊢ 𝐴 ∈ (1...𝐶)
Theorem	jm2.27dlem3 39106	Lemma for rmydioph 39109. Infer membership of the endpoint of a range. (Contributed by Stefan O'Rear, 11-Oct-2014.)
⊢ 𝐴 ∈ ℕ    ⇒   ⊢ 𝐴 ∈ (1...𝐴)
Theorem	jm2.27dlem4 39107	Lemma for rmydioph 39109. Infer ℕ-hood of large numbers. (Contributed by Stefan O'Rear, 11-Oct-2014.)
⊢ 𝐴 ∈ ℕ    &   ⊢ 𝐵 = (𝐴 + 1)    ⇒   ⊢ 𝐵 ∈ ℕ
Theorem	jm2.27dlem5 39108	Lemma for rmydioph 39109. Used with sselii 3888 to infer membership of midpoints of range; jm2.27dlem2 39105 is deprecated. (Contributed by Stefan O'Rear, 11-Oct-2014.)
⊢ 𝐵 = (𝐴 + 1)    &   ⊢ (1...𝐵) ⊆ (1...𝐶)    ⇒   ⊢ (1...𝐴) ⊆ (1...𝐶)
Theorem	rmydioph 39109	jm2.27 39103 restated in terms of Diophantine sets. (Contributed by Stefan O'Rear, 11-Oct-2014.) (Revised by Stefan O'Rear, 6-May-2015.)
⊢ {𝑎 ∈ (ℕ0 ↑𝑚 (1...3)) ∣ ((𝑎‘1) ∈ (ℤ≥‘2) ∧ (𝑎‘3) = ((𝑎‘1) Yrm (𝑎‘2)))} ∈ (Dioph‘3)
20.27.34  X and Y sequences 5: Diophantine representability of X, ^, _C
Theorem	rmxdiophlem 39110*	X can be expressed in terms of Y, so it is also Diophantine. (Contributed by Stefan O'Rear, 15-Oct-2014.)
⊢ ((𝐴 ∈ (ℤ≥‘2) ∧ 𝑁 ∈ ℕ0 ∧ 𝑋 ∈ ℕ0) → (𝑋 = (𝐴 Xrm 𝑁) ↔ ∃𝑦 ∈ ℕ0 (𝑦 = (𝐴 Yrm 𝑁) ∧ ((𝑋↑2) − (((𝐴↑2) − 1) · (𝑦↑2))) = 1)))
Theorem	rmxdioph 39111	X is a Diophantine function. (Contributed by Stefan O'Rear, 17-Oct-2014.)
⊢ {𝑎 ∈ (ℕ0 ↑𝑚 (1...3)) ∣ ((𝑎‘1) ∈ (ℤ≥‘2) ∧ (𝑎‘3) = ((𝑎‘1) Xrm (𝑎‘2)))} ∈ (Dioph‘3)
Theorem	jm3.1lem1 39112	Lemma for jm3.1 39115. (Contributed by Stefan O'Rear, 16-Oct-2014.)
⊢ (𝜑 → 𝐴 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝐾 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝑁 ∈ ℕ)    &   ⊢ (𝜑 → (𝐾 Yrm (𝑁 + 1)) ≤ 𝐴)    ⇒   ⊢ (𝜑 → (𝐾↑𝑁) < 𝐴)
Theorem	jm3.1lem2 39113	Lemma for jm3.1 39115. (Contributed by Stefan O'Rear, 16-Oct-2014.)
⊢ (𝜑 → 𝐴 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝐾 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝑁 ∈ ℕ)    &   ⊢ (𝜑 → (𝐾 Yrm (𝑁 + 1)) ≤ 𝐴)    ⇒   ⊢ (𝜑 → (𝐾↑𝑁) < ((((2 · 𝐴) · 𝐾) − (𝐾↑2)) − 1))
Theorem	jm3.1lem3 39114	Lemma for jm3.1 39115. (Contributed by Stefan O'Rear, 17-Oct-2014.)
⊢ (𝜑 → 𝐴 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝐾 ∈ (ℤ≥‘2))    &   ⊢ (𝜑 → 𝑁 ∈ ℕ)    &   ⊢ (𝜑 → (𝐾 Yrm (𝑁 + 1)) ≤ 𝐴)    ⇒   ⊢ (𝜑 → ((((2 · 𝐴) · 𝐾) − (𝐾↑2)) − 1) ∈ ℕ)
Theorem	jm3.1 39115	Diophantine expression for exponentiation. Lemma 3.1 of [JonesMatijasevic] p. 698. (Contributed by Stefan O'Rear, 16-Oct-2014.)
⊢ (((𝐴 ∈ (ℤ≥‘2) ∧ 𝐾 ∈ (ℤ≥‘2) ∧ 𝑁 ∈ ℕ) ∧ (𝐾 Yrm (𝑁 + 1)) ≤ 𝐴) → (𝐾↑𝑁) = (((𝐴 Xrm 𝑁) − ((𝐴 − 𝐾) · (𝐴 Yrm 𝑁))) mod ((((2 · 𝐴) · 𝐾) − (𝐾↑2)) − 1)))
Theorem	expdiophlem1 39116*	Lemma for expdioph 39118. Fully expanded expression for exponential. (Contributed by Stefan O'Rear, 17-Oct-2014.)
⊢ (𝐶 ∈ ℕ0 → (((𝐴 ∈ (ℤ≥‘2) ∧ 𝐵 ∈ ℕ) ∧ 𝐶 = (𝐴↑𝐵)) ↔ ∃𝑑 ∈ ℕ0 ∃𝑒 ∈ ℕ0 ∃𝑓 ∈ ℕ0 ((𝐴 ∈ (ℤ≥‘2) ∧ 𝐵 ∈ ℕ) ∧ ((𝐴 ∈ (ℤ≥‘2) ∧ 𝑑 = (𝐴 Yrm (𝐵 + 1))) ∧ ((𝑑 ∈ (ℤ≥‘2) ∧ 𝑒 = (𝑑 Yrm 𝐵)) ∧ ((𝑑 ∈ (ℤ≥‘2) ∧ 𝑓 = (𝑑 Xrm 𝐵)) ∧ (𝐶 < ((((2 · 𝑑) · 𝐴) − (𝐴↑2)) − 1) ∧ ((((2 · 𝑑) · 𝐴) − (𝐴↑2)) − 1) ∥ ((𝑓 − ((𝑑 − 𝐴) · 𝑒)) − 𝐶))))))))
Theorem	expdiophlem2 39117	Lemma for expdioph 39118. Exponentiation on a restricted domain is Diophantine. (Contributed by Stefan O'Rear, 17-Oct-2014.)
⊢ {𝑎 ∈ (ℕ0 ↑𝑚 (1...3)) ∣ (((𝑎‘1) ∈ (ℤ≥‘2) ∧ (𝑎‘2) ∈ ℕ) ∧ (𝑎‘3) = ((𝑎‘1)↑(𝑎‘2)))} ∈ (Dioph‘3)
Theorem	expdioph 39118	The exponential function is Diophantine. This result completes and encapsulates our development using Pell equation solution sequences and is sometimes regarded as Matiyasevich's theorem properly. (Contributed by Stefan O'Rear, 17-Oct-2014.)
⊢ {𝑎 ∈ (ℕ0 ↑𝑚 (1...3)) ∣ (𝑎‘3) = ((𝑎‘1)↑(𝑎‘2))} ∈ (Dioph‘3)
20.27.35  Uncategorized stuff not associated with a major project
Theorem	setindtr 39119*	Epsilon induction for sets contained in a transitive set. If we are allowed to assume Infinity, then all sets have a transitive closure and this reduces to setind 9025; however, this version is useful without Infinity. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ (∀𝑥(𝑥 ⊆ 𝐴 → 𝑥 ∈ 𝐴) → (∃𝑦(Tr 𝑦 ∧ 𝐵 ∈ 𝑦) → 𝐵 ∈ 𝐴))
Theorem	setindtrs 39120*	Epsilon induction scheme without Infinity. See comments at setindtr 39119. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ (∀𝑦 ∈ 𝑥 𝜓 → 𝜑)    &   ⊢ (𝑥 = 𝑦 → (𝜑 ↔ 𝜓))    &   ⊢ (𝑥 = 𝐵 → (𝜑 ↔ 𝜒))    ⇒   ⊢ (∃𝑧(Tr 𝑧 ∧ 𝐵 ∈ 𝑧) → 𝜒)
Theorem	dford3lem1 39121*	Lemma for dford3 39123. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ ((Tr 𝑁 ∧ ∀𝑦 ∈ 𝑁 Tr 𝑦) → ∀𝑏 ∈ 𝑁 (Tr 𝑏 ∧ ∀𝑦 ∈ 𝑏 Tr 𝑦))
Theorem	dford3lem2 39122*	Lemma for dford3 39123. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ ((Tr 𝑥 ∧ ∀𝑦 ∈ 𝑥 Tr 𝑦) → 𝑥 ∈ On)
Theorem	dford3 39123*	Ordinals are precisely the hereditarily transitive classes. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ (Ord 𝑁 ↔ (Tr 𝑁 ∧ ∀𝑥 ∈ 𝑁 Tr 𝑥))
Theorem	dford4 39124*	dford3 39123 expressed in primitives to demonstrate shortness. (Contributed by Stefan O'Rear, 28-Oct-2014.)
⊢ (Ord 𝑁 ↔ ∀𝑎∀𝑏∀𝑐((𝑎 ∈ 𝑁 ∧ 𝑏 ∈ 𝑎) → (𝑏 ∈ 𝑁 ∧ (𝑐 ∈ 𝑏 → 𝑐 ∈ 𝑎))))
Theorem	wopprc 39125	Unrelated: Wiener pairs treat proper classes symmetrically. (Contributed by Stefan O'Rear, 19-Sep-2014.)
⊢ ((𝐴 ∈ V ∧ 𝐵 ∈ V) ↔ ¬ 1o ∈ {{{𝐴}, ∅}, {{𝐵}}})
Theorem	rpnnen3lem 39126*	Lemma for rpnnen3 39127. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ (((𝑎 ∈ ℝ ∧ 𝑏 ∈ ℝ) ∧ 𝑎 < 𝑏) → {𝑐 ∈ ℚ ∣ 𝑐 < 𝑎} ≠ {𝑐 ∈ ℚ ∣ 𝑐 < 𝑏})
Theorem	rpnnen3 39127	Dedekind cut injection of ℝ into 𝒫 ℚ. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ ℝ ≼ 𝒫 ℚ
20.27.36  More equivalents of the Axiom of Choice
Theorem	axac10 39128	Characterization of choice similar to dffin1-5 9659. (Contributed by Stefan O'Rear, 6-Jan-2015.)
⊢ ( ≈ “ On) = V
Theorem	harinf 39129	The Hartogs number of an infinite set is at least ω. MOVABLE (Contributed by Stefan O'Rear, 10-Jul-2015.)
⊢ ((𝑆 ∈ 𝑉 ∧ ¬ 𝑆 ∈ Fin) → ω ⊆ (har‘𝑆))
Theorem	wdom2d2 39130*	Deduction for weak dominance by a Cartesian product. MOVABLE (Contributed by Stefan O'Rear, 10-Jul-2015.)
⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → 𝐵 ∈ 𝑊)    &   ⊢ (𝜑 → 𝐶 ∈ 𝑋)    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴) → ∃𝑦 ∈ 𝐵 ∃𝑧 ∈ 𝐶 𝑥 = 𝑋)    ⇒   ⊢ (𝜑 → 𝐴 ≼* (𝐵 × 𝐶))
Theorem	ttac 39131	Tarski's theorem about choice: infxpidm 9833 is equivalent to ax-ac 9730. (Contributed by Stefan O'Rear, 4-Nov-2014.) (Proof shortened by Stefan O'Rear, 10-Jul-2015.)
⊢ (CHOICE ↔ ∀𝑐(ω ≼ 𝑐 → (𝑐 × 𝑐) ≈ 𝑐))
Theorem	pw2f1ocnv 39132*	Define a bijection between characteristic functions and subsets. EDITORIAL: extracted from pw2en 8474, which can be easily reproved in terms of this. (Contributed by Stefan O'Rear, 18-Jan-2015.) (Revised by Stefan O'Rear, 9-Jul-2015.)
⊢ 𝐹 = (𝑥 ∈ (2o ↑𝑚 𝐴) ↦ (◡𝑥 “ {1o}))    ⇒   ⊢ (𝐴 ∈ 𝑉 → (𝐹:(2o ↑𝑚 𝐴)–1-1-onto→𝒫 𝐴 ∧ ◡𝐹 = (𝑦 ∈ 𝒫 𝐴 ↦ (𝑧 ∈ 𝐴 ↦ if(𝑧 ∈ 𝑦, 1o, ∅)))))
Theorem	pw2f1o2 39133*	Define a bijection between characteristic functions and subsets. EDITORIAL: extracted from pw2en 8474, which can be easily reproved in terms of this. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = (𝑥 ∈ (2o ↑𝑚 𝐴) ↦ (◡𝑥 “ {1o}))    ⇒   ⊢ (𝐴 ∈ 𝑉 → 𝐹:(2o ↑𝑚 𝐴)–1-1-onto→𝒫 𝐴)
Theorem	pw2f1o2val 39134*	Function value of the pw2f1o2 39133 bijection. (Contributed by Stefan O'Rear, 18-Jan-2015.) (Revised by Stefan O'Rear, 6-May-2015.)
⊢ 𝐹 = (𝑥 ∈ (2o ↑𝑚 𝐴) ↦ (◡𝑥 “ {1o}))    ⇒   ⊢ (𝑋 ∈ (2o ↑𝑚 𝐴) → (𝐹‘𝑋) = (◡𝑋 “ {1o}))
Theorem	pw2f1o2val2 39135*	Membership in a mapped set under the pw2f1o2 39133 bijection. (Contributed by Stefan O'Rear, 18-Jan-2015.) (Revised by Stefan O'Rear, 6-May-2015.)
⊢ 𝐹 = (𝑥 ∈ (2o ↑𝑚 𝐴) ↦ (◡𝑥 “ {1o}))    ⇒   ⊢ ((𝑋 ∈ (2o ↑𝑚 𝐴) ∧ 𝑌 ∈ 𝐴) → (𝑌 ∈ (𝐹‘𝑋) ↔ (𝑋‘𝑌) = 1o))
Theorem	soeq12d 39136	Equality deduction for total orderings. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝜑 → 𝑅 = 𝑆)    &   ⊢ (𝜑 → 𝐴 = 𝐵)    ⇒   ⊢ (𝜑 → (𝑅 Or 𝐴 ↔ 𝑆 Or 𝐵))
Theorem	freq12d 39137	Equality deduction for founded relations. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝜑 → 𝑅 = 𝑆)    &   ⊢ (𝜑 → 𝐴 = 𝐵)    ⇒   ⊢ (𝜑 → (𝑅 Fr 𝐴 ↔ 𝑆 Fr 𝐵))
Theorem	weeq12d 39138	Equality deduction for well-orders. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝜑 → 𝑅 = 𝑆)    &   ⊢ (𝜑 → 𝐴 = 𝐵)    ⇒   ⊢ (𝜑 → (𝑅 We 𝐴 ↔ 𝑆 We 𝐵))
Theorem	limsuc2 39139	Limit ordinals in the sense inclusive of zero contain all successors of their members. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ ((Ord 𝐴 ∧ 𝐴 = ∪ 𝐴) → (𝐵 ∈ 𝐴 ↔ suc 𝐵 ∈ 𝐴))
Theorem	wepwsolem 39140*	Transfer an ordering on characteristic functions by isomorphism to the power set. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ∃𝑧 ∈ 𝐴 ((𝑧 ∈ 𝑦 ∧ ¬ 𝑧 ∈ 𝑥) ∧ ∀𝑤 ∈ 𝐴 (𝑤𝑅𝑧 → (𝑤 ∈ 𝑥 ↔ 𝑤 ∈ 𝑦)))}    &   ⊢ 𝑈 = {⟨𝑥, 𝑦⟩ ∣ ∃𝑧 ∈ 𝐴 ((𝑥‘𝑧) E (𝑦‘𝑧) ∧ ∀𝑤 ∈ 𝐴 (𝑤𝑅𝑧 → (𝑥‘𝑤) = (𝑦‘𝑤)))}    &   ⊢ 𝐹 = (𝑎 ∈ (2o ↑𝑚 𝐴) ↦ (◡𝑎 “ {1o}))    ⇒   ⊢ (𝐴 ∈ V → 𝐹 Isom 𝑈, 𝑇 ((2o ↑𝑚 𝐴), 𝒫 𝐴))
Theorem	wepwso 39141*	A well-ordering induces a strict ordering on the power set. EDITORIAL: when well-orderings are set like, this can be strengthened to remove 𝐴 ∈ 𝑉. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ∃𝑧 ∈ 𝐴 ((𝑧 ∈ 𝑦 ∧ ¬ 𝑧 ∈ 𝑥) ∧ ∀𝑤 ∈ 𝐴 (𝑤𝑅𝑧 → (𝑤 ∈ 𝑥 ↔ 𝑤 ∈ 𝑦)))}    ⇒   ⊢ ((𝐴 ∈ 𝑉 ∧ 𝑅 We 𝐴) → 𝑇 Or 𝒫 𝐴)
Theorem	dnnumch1 39142*	Define an enumeration of a set from a choice function; second part, it restricts to a bijection. EDITORIAL: overlaps dfac8a 9305. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = recs((𝑧 ∈ V ↦ (𝐺‘(𝐴 ∖ ran 𝑧))))    &   ⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → ∀𝑦 ∈ 𝒫 𝐴(𝑦 ≠ ∅ → (𝐺‘𝑦) ∈ 𝑦))    ⇒   ⊢ (𝜑 → ∃𝑥 ∈ On (𝐹 ↾ 𝑥):𝑥–1-1-onto→𝐴)
Theorem	dnnumch2 39143*	Define an enumeration (weak dominance version) of a set from a choice function. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = recs((𝑧 ∈ V ↦ (𝐺‘(𝐴 ∖ ran 𝑧))))    &   ⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → ∀𝑦 ∈ 𝒫 𝐴(𝑦 ≠ ∅ → (𝐺‘𝑦) ∈ 𝑦))    ⇒   ⊢ (𝜑 → 𝐴 ⊆ ran 𝐹)
Theorem	dnnumch3lem 39144*	Value of the ordinal injection function. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = recs((𝑧 ∈ V ↦ (𝐺‘(𝐴 ∖ ran 𝑧))))    &   ⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → ∀𝑦 ∈ 𝒫 𝐴(𝑦 ≠ ∅ → (𝐺‘𝑦) ∈ 𝑦))    ⇒   ⊢ ((𝜑 ∧ 𝑤 ∈ 𝐴) → ((𝑥 ∈ 𝐴 ↦ ∩ (◡𝐹 “ {𝑥}))‘𝑤) = ∩ (◡𝐹 “ {𝑤}))
Theorem	dnnumch3 39145*	Define an injection from a set into the ordinals using a choice function. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = recs((𝑧 ∈ V ↦ (𝐺‘(𝐴 ∖ ran 𝑧))))    &   ⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → ∀𝑦 ∈ 𝒫 𝐴(𝑦 ≠ ∅ → (𝐺‘𝑦) ∈ 𝑦))    ⇒   ⊢ (𝜑 → (𝑥 ∈ 𝐴 ↦ ∩ (◡𝐹 “ {𝑥})):𝐴–1-1→On)
Theorem	dnwech 39146*	Define a well-ordering from a choice function. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐹 = recs((𝑧 ∈ V ↦ (𝐺‘(𝐴 ∖ ran 𝑧))))    &   ⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → ∀𝑦 ∈ 𝒫 𝐴(𝑦 ≠ ∅ → (𝐺‘𝑦) ∈ 𝑦))    &   ⊢ 𝐻 = {⟨𝑣, 𝑤⟩ ∣ ∩ (◡𝐹 “ {𝑣}) ∈ ∩ (◡𝐹 “ {𝑤})}    ⇒   ⊢ (𝜑 → 𝐻 We 𝐴)
Theorem	fnwe2val 39147*	Lemma for fnwe2 39151. Substitute variables. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝑧 = (𝐹‘𝑥) → 𝑆 = 𝑈)    &   ⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ((𝐹‘𝑥)𝑅(𝐹‘𝑦) ∨ ((𝐹‘𝑥) = (𝐹‘𝑦) ∧ 𝑥𝑈𝑦))}    ⇒   ⊢ (𝑎𝑇𝑏 ↔ ((𝐹‘𝑎)𝑅(𝐹‘𝑏) ∨ ((𝐹‘𝑎) = (𝐹‘𝑏) ∧ 𝑎⦋(𝐹‘𝑎) / 𝑧⦌𝑆𝑏)))
Theorem	fnwe2lem1 39148*	Lemma for fnwe2 39151. Substitution in well-ordering hypothesis. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝑧 = (𝐹‘𝑥) → 𝑆 = 𝑈)    &   ⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ((𝐹‘𝑥)𝑅(𝐹‘𝑦) ∨ ((𝐹‘𝑥) = (𝐹‘𝑦) ∧ 𝑥𝑈𝑦))}    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴) → 𝑈 We {𝑦 ∈ 𝐴 ∣ (𝐹‘𝑦) = (𝐹‘𝑥)})    ⇒   ⊢ ((𝜑 ∧ 𝑎 ∈ 𝐴) → ⦋(𝐹‘𝑎) / 𝑧⦌𝑆 We {𝑦 ∈ 𝐴 ∣ (𝐹‘𝑦) = (𝐹‘𝑎)})
Theorem	fnwe2lem2 39149*	Lemma for fnwe2 39151. An element which is in a minimal fiber and minimal within its fiber is minimal globally; thus 𝑇 is well-founded. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝑧 = (𝐹‘𝑥) → 𝑆 = 𝑈)    &   ⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ((𝐹‘𝑥)𝑅(𝐹‘𝑦) ∨ ((𝐹‘𝑥) = (𝐹‘𝑦) ∧ 𝑥𝑈𝑦))}    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴) → 𝑈 We {𝑦 ∈ 𝐴 ∣ (𝐹‘𝑦) = (𝐹‘𝑥)})    &   ⊢ (𝜑 → (𝐹 ↾ 𝐴):𝐴⟶𝐵)    &   ⊢ (𝜑 → 𝑅 We 𝐵)    &   ⊢ (𝜑 → 𝑎 ⊆ 𝐴)    &   ⊢ (𝜑 → 𝑎 ≠ ∅)    ⇒   ⊢ (𝜑 → ∃𝑏 ∈ 𝑎 ∀𝑐 ∈ 𝑎 ¬ 𝑐𝑇𝑏)
Theorem	fnwe2lem3 39150*	Lemma for fnwe2 39151. Trichotomy. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝑧 = (𝐹‘𝑥) → 𝑆 = 𝑈)    &   ⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ((𝐹‘𝑥)𝑅(𝐹‘𝑦) ∨ ((𝐹‘𝑥) = (𝐹‘𝑦) ∧ 𝑥𝑈𝑦))}    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴) → 𝑈 We {𝑦 ∈ 𝐴 ∣ (𝐹‘𝑦) = (𝐹‘𝑥)})    &   ⊢ (𝜑 → (𝐹 ↾ 𝐴):𝐴⟶𝐵)    &   ⊢ (𝜑 → 𝑅 We 𝐵)    &   ⊢ (𝜑 → 𝑎 ∈ 𝐴)    &   ⊢ (𝜑 → 𝑏 ∈ 𝐴)    ⇒   ⊢ (𝜑 → (𝑎𝑇𝑏 ∨ 𝑎 = 𝑏 ∨ 𝑏𝑇𝑎))
Theorem	fnwe2 39151*	A well-ordering can be constructed on a partitioned set by patching together well-orderings on each partition using a well-ordering on the partitions themselves. Similar to fnwe 7682 but does not require the within-partition ordering to be globally well. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ (𝑧 = (𝐹‘𝑥) → 𝑆 = 𝑈)    &   ⊢ 𝑇 = {⟨𝑥, 𝑦⟩ ∣ ((𝐹‘𝑥)𝑅(𝐹‘𝑦) ∨ ((𝐹‘𝑥) = (𝐹‘𝑦) ∧ 𝑥𝑈𝑦))}    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴) → 𝑈 We {𝑦 ∈ 𝐴 ∣ (𝐹‘𝑦) = (𝐹‘𝑥)})    &   ⊢ (𝜑 → (𝐹 ↾ 𝐴):𝐴⟶𝐵)    &   ⊢ (𝜑 → 𝑅 We 𝐵)    ⇒   ⊢ (𝜑 → 𝑇 We 𝐴)
Theorem	aomclem1 39152*	Lemma for dfac11 39160. This is the beginning of the proof that multiple choice is equivalent to choice. Our goal is to construct, by transfinite recursion, a well-ordering of (𝑅1‘𝐴). In what follows, 𝐴 is the index of the rank we wish to well-order, 𝑧 is the collection of well-orderings constructed so far, dom 𝑧 is the set of ordinal indices of constructed ranks i.e. the next rank to construct, and 𝑦 is a postulated multiple-choice function. Successor case 1, define a simple ordering from the well-ordered predecessor. (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ (𝜑 → dom 𝑧 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 = suc ∪ dom 𝑧)    &   ⊢ (𝜑 → ∀𝑎 ∈ dom 𝑧(𝑧‘𝑎) We (𝑅1‘𝑎))    ⇒   ⊢ (𝜑 → 𝐵 Or (𝑅1‘dom 𝑧))
Theorem	aomclem2 39153*	Lemma for dfac11 39160. Successor case 2, a choice function for subsets of (𝑅1‘dom 𝑧). (Contributed by Stefan O'Rear, 18-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ 𝐶 = (𝑎 ∈ V ↦ sup((𝑦‘𝑎), (𝑅1‘dom 𝑧), 𝐵))    &   ⊢ (𝜑 → dom 𝑧 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 = suc ∪ dom 𝑧)    &   ⊢ (𝜑 → ∀𝑎 ∈ dom 𝑧(𝑧‘𝑎) We (𝑅1‘𝑎))    &   ⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 ⊆ 𝐴)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘dom 𝑧)(𝑎 ≠ ∅ → (𝐶‘𝑎) ∈ 𝑎))
Theorem	aomclem3 39154*	Lemma for dfac11 39160. Successor case 3, our required well-ordering. (Contributed by Stefan O'Rear, 19-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ 𝐶 = (𝑎 ∈ V ↦ sup((𝑦‘𝑎), (𝑅1‘dom 𝑧), 𝐵))    &   ⊢ 𝐷 = recs((𝑎 ∈ V ↦ (𝐶‘((𝑅1‘dom 𝑧) ∖ ran 𝑎))))    &   ⊢ 𝐸 = {⟨𝑎, 𝑏⟩ ∣ ∩ (◡𝐷 “ {𝑎}) ∈ ∩ (◡𝐷 “ {𝑏})}    &   ⊢ (𝜑 → dom 𝑧 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 = suc ∪ dom 𝑧)    &   ⊢ (𝜑 → ∀𝑎 ∈ dom 𝑧(𝑧‘𝑎) We (𝑅1‘𝑎))    &   ⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 ⊆ 𝐴)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → 𝐸 We (𝑅1‘dom 𝑧))
Theorem	aomclem4 39155*	Lemma for dfac11 39160. Limit case. Patch together well-orderings constructed so far using fnwe2 39151 to cover the limit rank. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ 𝐹 = {⟨𝑎, 𝑏⟩ ∣ ((rank‘𝑎) E (rank‘𝑏) ∨ ((rank‘𝑎) = (rank‘𝑏) ∧ 𝑎(𝑧‘suc (rank‘𝑎))𝑏))}    &   ⊢ (𝜑 → dom 𝑧 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 = ∪ dom 𝑧)    &   ⊢ (𝜑 → ∀𝑎 ∈ dom 𝑧(𝑧‘𝑎) We (𝑅1‘𝑎))    ⇒   ⊢ (𝜑 → 𝐹 We (𝑅1‘dom 𝑧))
Theorem	aomclem5 39156*	Lemma for dfac11 39160. Combine the successor case with the limit case. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ 𝐶 = (𝑎 ∈ V ↦ sup((𝑦‘𝑎), (𝑅1‘dom 𝑧), 𝐵))    &   ⊢ 𝐷 = recs((𝑎 ∈ V ↦ (𝐶‘((𝑅1‘dom 𝑧) ∖ ran 𝑎))))    &   ⊢ 𝐸 = {⟨𝑎, 𝑏⟩ ∣ ∩ (◡𝐷 “ {𝑎}) ∈ ∩ (◡𝐷 “ {𝑏})}    &   ⊢ 𝐹 = {⟨𝑎, 𝑏⟩ ∣ ((rank‘𝑎) E (rank‘𝑏) ∨ ((rank‘𝑎) = (rank‘𝑏) ∧ 𝑎(𝑧‘suc (rank‘𝑎))𝑏))}    &   ⊢ 𝐺 = (if(dom 𝑧 = ∪ dom 𝑧, 𝐹, 𝐸) ∩ ((𝑅1‘dom 𝑧) × (𝑅1‘dom 𝑧)))    &   ⊢ (𝜑 → dom 𝑧 ∈ On)    &   ⊢ (𝜑 → ∀𝑎 ∈ dom 𝑧(𝑧‘𝑎) We (𝑅1‘𝑎))    &   ⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → dom 𝑧 ⊆ 𝐴)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → 𝐺 We (𝑅1‘dom 𝑧))
Theorem	aomclem6 39157*	Lemma for dfac11 39160. Transfinite induction, close over 𝑧. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ 𝐶 = (𝑎 ∈ V ↦ sup((𝑦‘𝑎), (𝑅1‘dom 𝑧), 𝐵))    &   ⊢ 𝐷 = recs((𝑎 ∈ V ↦ (𝐶‘((𝑅1‘dom 𝑧) ∖ ran 𝑎))))    &   ⊢ 𝐸 = {⟨𝑎, 𝑏⟩ ∣ ∩ (◡𝐷 “ {𝑎}) ∈ ∩ (◡𝐷 “ {𝑏})}    &   ⊢ 𝐹 = {⟨𝑎, 𝑏⟩ ∣ ((rank‘𝑎) E (rank‘𝑏) ∨ ((rank‘𝑎) = (rank‘𝑏) ∧ 𝑎(𝑧‘suc (rank‘𝑎))𝑏))}    &   ⊢ 𝐺 = (if(dom 𝑧 = ∪ dom 𝑧, 𝐹, 𝐸) ∩ ((𝑅1‘dom 𝑧) × (𝑅1‘dom 𝑧)))    &   ⊢ 𝐻 = recs((𝑧 ∈ V ↦ 𝐺))    &   ⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → (𝐻‘𝐴) We (𝑅1‘𝐴))
Theorem	aomclem7 39158*	Lemma for dfac11 39160. (𝑅1‘𝐴) is well-orderable. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ 𝐵 = {⟨𝑎, 𝑏⟩ ∣ ∃𝑐 ∈ (𝑅1‘∪ dom 𝑧)((𝑐 ∈ 𝑏 ∧ ¬ 𝑐 ∈ 𝑎) ∧ ∀𝑑 ∈ (𝑅1‘∪ dom 𝑧)(𝑑(𝑧‘∪ dom 𝑧)𝑐 → (𝑑 ∈ 𝑎 ↔ 𝑑 ∈ 𝑏)))}    &   ⊢ 𝐶 = (𝑎 ∈ V ↦ sup((𝑦‘𝑎), (𝑅1‘dom 𝑧), 𝐵))    &   ⊢ 𝐷 = recs((𝑎 ∈ V ↦ (𝐶‘((𝑅1‘dom 𝑧) ∖ ran 𝑎))))    &   ⊢ 𝐸 = {⟨𝑎, 𝑏⟩ ∣ ∩ (◡𝐷 “ {𝑎}) ∈ ∩ (◡𝐷 “ {𝑏})}    &   ⊢ 𝐹 = {⟨𝑎, 𝑏⟩ ∣ ((rank‘𝑎) E (rank‘𝑏) ∨ ((rank‘𝑎) = (rank‘𝑏) ∧ 𝑎(𝑧‘suc (rank‘𝑎))𝑏))}    &   ⊢ 𝐺 = (if(dom 𝑧 = ∪ dom 𝑧, 𝐹, 𝐸) ∩ ((𝑅1‘dom 𝑧) × (𝑅1‘dom 𝑧)))    &   ⊢ 𝐻 = recs((𝑧 ∈ V ↦ 𝐺))    &   ⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → ∃𝑏 𝑏 We (𝑅1‘𝐴))
Theorem	aomclem8 39159*	Lemma for dfac11 39160. Perform variable substitutions. This is the most we can say without invoking regularity. (Contributed by Stefan O'Rear, 20-Jan-2015.)
⊢ (𝜑 → 𝐴 ∈ On)    &   ⊢ (𝜑 → ∀𝑎 ∈ 𝒫 (𝑅1‘𝐴)(𝑎 ≠ ∅ → (𝑦‘𝑎) ∈ ((𝒫 𝑎 ∩ Fin) ∖ {∅})))    ⇒   ⊢ (𝜑 → ∃𝑏 𝑏 We (𝑅1‘𝐴))
Theorem	dfac11 39160*	The right-hand side of this theorem (compare with ac4 9746), sometimes known as the "axiom of multiple choice", is a choice equivalent. Curiously, this statement cannot be proved without ax-reg 8905, despite not mentioning the cumulative hierarchy in any way as most consequences of regularity do. This is definition (MC) of [Schechter] p. 141. EDITORIAL: the proof is not original with me of course but I lost my reference sometime after writing it. A multiple choice function allows any total order to be extended to a choice function, which in turn defines a well-ordering. Since a well-ordering on a set defines a simple ordering of the power set, this allows the trivial well-ordering of the empty set to be transfinitely bootstrapped up the cumulative hierarchy to any desired level. (Contributed by Stefan O'Rear, 20-Jan-2015.) (Revised by Stefan O'Rear, 1-Jun-2015.)
⊢ (CHOICE ↔ ∀𝑥∃𝑓∀𝑧 ∈ 𝑥 (𝑧 ≠ ∅ → (𝑓‘𝑧) ∈ ((𝒫 𝑧 ∩ Fin) ∖ {∅})))
Theorem	kelac1 39161*	Kelley's choice, basic form: if a collection of sets can be cast as closed sets in the factors of a topology, and there is a definable element in each topology (which need not be in the closed set - if it were this would be trivial), then compactness (via finite intersection) guarantees that the final product is nonempty. (Contributed by Stefan O'Rear, 22-Feb-2015.)
⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝑆 ≠ ∅)    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝐽 ∈ Top)    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝐶 ∈ (Clsd‘𝐽))    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝐵:𝑆–1-1-onto→𝐶)    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝑈 ∈ ∪ 𝐽)    &   ⊢ (𝜑 → (∏t‘(𝑥 ∈ 𝐼 ↦ 𝐽)) ∈ Comp)    ⇒   ⊢ (𝜑 → X𝑥 ∈ 𝐼 𝑆 ≠ ∅)
Theorem	kelac2lem 39162	Lemma for kelac2 39163 and dfac21 39164: knob topologies are compact. (Contributed by Stefan O'Rear, 22-Feb-2015.)
⊢ (𝑆 ∈ 𝑉 → (topGen‘{𝑆, {𝒫 ∪ 𝑆}}) ∈ Comp)
Theorem	kelac2 39163*	Kelley's choice, most common form: compactness of a product of knob topologies recovers choice. (Contributed by Stefan O'Rear, 22-Feb-2015.)
⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝑆 ∈ 𝑉)    &   ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐼) → 𝑆 ≠ ∅)    &   ⊢ (𝜑 → (∏t‘(𝑥 ∈ 𝐼 ↦ (topGen‘{𝑆, {𝒫 ∪ 𝑆}}))) ∈ Comp)    ⇒   ⊢ (𝜑 → X𝑥 ∈ 𝐼 𝑆 ≠ ∅)
Theorem	dfac21 39164	Tychonoff's theorem is a choice equivalent. Definition AC21 of Schechter p. 461. (Contributed by Stefan O'Rear, 22-Feb-2015.) (Revised by Mario Carneiro, 27-Aug-2015.)
⊢ (CHOICE ↔ ∀𝑓(𝑓:dom 𝑓⟶Comp → (∏t‘𝑓) ∈ Comp))
20.27.37  Finitely generated left modules
Syntax	clfig 39165	Extend class notation with the class of finitely generated left modules.
class LFinGen
Definition	df-lfig 39166	Define the class of finitely generated left modules. Finite generation of subspaces can be intepreted using ↾s. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ LFinGen = {𝑤 ∈ LMod ∣ (Base‘𝑤) ∈ ((LSpan‘𝑤) “ (𝒫 (Base‘𝑤) ∩ Fin))}
Theorem	islmodfg 39167*	Property of a finitely generated left module. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ 𝐵 = (Base‘𝑊)    &   ⊢ 𝑁 = (LSpan‘𝑊)    ⇒   ⊢ (𝑊 ∈ LMod → (𝑊 ∈ LFinGen ↔ ∃𝑏 ∈ 𝒫 𝐵(𝑏 ∈ Fin ∧ (𝑁‘𝑏) = 𝐵)))
Theorem	islssfg 39168*	Property of a finitely generated left (sub)module. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ 𝑋 = (𝑊 ↾s 𝑈)    &   ⊢ 𝑆 = (LSubSp‘𝑊)    &   ⊢ 𝑁 = (LSpan‘𝑊)    ⇒   ⊢ ((𝑊 ∈ LMod ∧ 𝑈 ∈ 𝑆) → (𝑋 ∈ LFinGen ↔ ∃𝑏 ∈ 𝒫 𝑈(𝑏 ∈ Fin ∧ (𝑁‘𝑏) = 𝑈)))
Theorem	islssfg2 39169*	Property of a finitely generated left (sub)module, with a relaxed constraint on the spanning vectors. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑋 = (𝑊 ↾s 𝑈)    &   ⊢ 𝑆 = (LSubSp‘𝑊)    &   ⊢ 𝑁 = (LSpan‘𝑊)    &   ⊢ 𝐵 = (Base‘𝑊)    ⇒   ⊢ ((𝑊 ∈ LMod ∧ 𝑈 ∈ 𝑆) → (𝑋 ∈ LFinGen ↔ ∃𝑏 ∈ (𝒫 𝐵 ∩ Fin)(𝑁‘𝑏) = 𝑈))
Theorem	islssfgi 39170	Finitely spanned subspaces are finitely generated. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑁 = (LSpan‘𝑊)    &   ⊢ 𝑉 = (Base‘𝑊)    &   ⊢ 𝑋 = (𝑊 ↾s (𝑁‘𝐵))    ⇒   ⊢ ((𝑊 ∈ LMod ∧ 𝐵 ⊆ 𝑉 ∧ 𝐵 ∈ Fin) → 𝑋 ∈ LFinGen)
Theorem	fglmod 39171	Finitely generated left modules are left modules. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ (𝑀 ∈ LFinGen → 𝑀 ∈ LMod)
Theorem	lsmfgcl 39172	The sum of two finitely generated submodules is finitely generated. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑈 = (LSubSp‘𝑊)    &   ⊢ ⊕ = (LSSum‘𝑊)    &   ⊢ 𝐷 = (𝑊 ↾s 𝐴)    &   ⊢ 𝐸 = (𝑊 ↾s 𝐵)    &   ⊢ 𝐹 = (𝑊 ↾s (𝐴 ⊕ 𝐵))    &   ⊢ (𝜑 → 𝑊 ∈ LMod)    &   ⊢ (𝜑 → 𝐴 ∈ 𝑈)    &   ⊢ (𝜑 → 𝐵 ∈ 𝑈)    &   ⊢ (𝜑 → 𝐷 ∈ LFinGen)    &   ⊢ (𝜑 → 𝐸 ∈ LFinGen)    ⇒   ⊢ (𝜑 → 𝐹 ∈ LFinGen)
20.27.38  Noetherian left modules I
Syntax	clnm 39173	Extend class notation with the class of Noetherian left modules.
class LNoeM
Definition	df-lnm 39174*	A left-module is Noetherian iff it is hereditarily finitely generated. (Contributed by Stefan O'Rear, 12-Dec-2014.)
⊢ LNoeM = {𝑤 ∈ LMod ∣ ∀𝑖 ∈ (LSubSp‘𝑤)(𝑤 ↾s 𝑖) ∈ LFinGen}
Theorem	islnm 39175*	Property of being a Noetherian left module. (Contributed by Stefan O'Rear, 12-Dec-2014.)
⊢ 𝑆 = (LSubSp‘𝑀)    ⇒   ⊢ (𝑀 ∈ LNoeM ↔ (𝑀 ∈ LMod ∧ ∀𝑖 ∈ 𝑆 (𝑀 ↾s 𝑖) ∈ LFinGen))
Theorem	islnm2 39176*	Property of being a Noetherian left module with finite generation expanded in terms of spans. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝐵 = (Base‘𝑀)    &   ⊢ 𝑆 = (LSubSp‘𝑀)    &   ⊢ 𝑁 = (LSpan‘𝑀)    ⇒   ⊢ (𝑀 ∈ LNoeM ↔ (𝑀 ∈ LMod ∧ ∀𝑖 ∈ 𝑆 ∃𝑔 ∈ (𝒫 𝐵 ∩ Fin)𝑖 = (𝑁‘𝑔)))
Theorem	lnmlmod 39177	A Noetherian left module is a left module. (Contributed by Stefan O'Rear, 12-Dec-2014.)
⊢ (𝑀 ∈ LNoeM → 𝑀 ∈ LMod)
Theorem	lnmlssfg 39178	A submodule of Noetherian module is finitely generated. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ 𝑆 = (LSubSp‘𝑀)    &   ⊢ 𝑅 = (𝑀 ↾s 𝑈)    ⇒   ⊢ ((𝑀 ∈ LNoeM ∧ 𝑈 ∈ 𝑆) → 𝑅 ∈ LFinGen)
Theorem	lnmlsslnm 39179	All submodules of a Noetherian module are Noetherian. (Contributed by Stefan O'Rear, 1-Jan-2015.)
⊢ 𝑆 = (LSubSp‘𝑀)    &   ⊢ 𝑅 = (𝑀 ↾s 𝑈)    ⇒   ⊢ ((𝑀 ∈ LNoeM ∧ 𝑈 ∈ 𝑆) → 𝑅 ∈ LNoeM)
Theorem	lnmfg 39180	A Noetherian left module is finitely generated. (Contributed by Stefan O'Rear, 12-Dec-2014.)
⊢ (𝑀 ∈ LNoeM → 𝑀 ∈ LFinGen)
Theorem	kercvrlsm 39181	The domain of a linear function is the subspace sum of the kernel and any subspace which covers the range. (Contributed by Stefan O'Rear, 24-Jan-2015.) (Revised by Stefan O'Rear, 6-May-2015.)
⊢ 𝑈 = (LSubSp‘𝑆)    &   ⊢ ⊕ = (LSSum‘𝑆)    &   ⊢ 0 = (0g‘𝑇)    &   ⊢ 𝐾 = (◡𝐹 “ { 0 })    &   ⊢ 𝐵 = (Base‘𝑆)    &   ⊢ (𝜑 → 𝐹 ∈ (𝑆 LMHom 𝑇))    &   ⊢ (𝜑 → 𝐷 ∈ 𝑈)    &   ⊢ (𝜑 → (𝐹 “ 𝐷) = ran 𝐹)    ⇒   ⊢ (𝜑 → (𝐾 ⊕ 𝐷) = 𝐵)
Theorem	lmhmfgima 39182	A homomorphism maps finitely generated submodules to finitely generated submodules. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑌 = (𝑇 ↾s (𝐹 “ 𝐴))    &   ⊢ 𝑋 = (𝑆 ↾s 𝐴)    &   ⊢ 𝑈 = (LSubSp‘𝑆)    &   ⊢ (𝜑 → 𝑋 ∈ LFinGen)    &   ⊢ (𝜑 → 𝐴 ∈ 𝑈)    &   ⊢ (𝜑 → 𝐹 ∈ (𝑆 LMHom 𝑇))    ⇒   ⊢ (𝜑 → 𝑌 ∈ LFinGen)
Theorem	lnmepi 39183	Epimorphic images of Noetherian modules are Noetherian. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝐵 = (Base‘𝑇)    ⇒   ⊢ ((𝐹 ∈ (𝑆 LMHom 𝑇) ∧ 𝑆 ∈ LNoeM ∧ ran 𝐹 = 𝐵) → 𝑇 ∈ LNoeM)
Theorem	lmhmfgsplit 39184	If the kernel and range of a homomorphism of left modules are finitely generated, then so is the domain. (Contributed by Stefan O'Rear, 1-Jan-2015.) (Revised by Stefan O'Rear, 6-May-2015.)
⊢ 0 = (0g‘𝑇)    &   ⊢ 𝐾 = (◡𝐹 “ { 0 })    &   ⊢ 𝑈 = (𝑆 ↾s 𝐾)    &   ⊢ 𝑉 = (𝑇 ↾s ran 𝐹)    ⇒   ⊢ ((𝐹 ∈ (𝑆 LMHom 𝑇) ∧ 𝑈 ∈ LFinGen ∧ 𝑉 ∈ LFinGen) → 𝑆 ∈ LFinGen)
Theorem	lmhmlnmsplit 39185	If the kernel and range of a homomorphism of left modules are Noetherian, then so is the domain. (Contributed by Stefan O'Rear, 1-Jan-2015.) (Revised by Stefan O'Rear, 12-Jun-2015.)
⊢ 0 = (0g‘𝑇)    &   ⊢ 𝐾 = (◡𝐹 “ { 0 })    &   ⊢ 𝑈 = (𝑆 ↾s 𝐾)    &   ⊢ 𝑉 = (𝑇 ↾s ran 𝐹)    ⇒   ⊢ ((𝐹 ∈ (𝑆 LMHom 𝑇) ∧ 𝑈 ∈ LNoeM ∧ 𝑉 ∈ LNoeM) → 𝑆 ∈ LNoeM)
Theorem	lnmlmic 39186	Noetherian is an invariant property of modules. (Contributed by Stefan O'Rear, 25-Jan-2015.)
⊢ (𝑅 ≃𝑚 𝑆 → (𝑅 ∈ LNoeM ↔ 𝑆 ∈ LNoeM))
20.27.39  Addenda for structure powers
Theorem	pwssplit4 39187*	Splitting for structure powers 4: maps isomorphically onto the other half. (Contributed by Stefan O'Rear, 25-Jan-2015.)
⊢ 𝐸 = (𝑅 ↑s (𝐴 ∪ 𝐵))    &   ⊢ 𝐺 = (Base‘𝐸)    &   ⊢ 0 = (0g‘𝑅)    &   ⊢ 𝐾 = {𝑦 ∈ 𝐺 ∣ (𝑦 ↾ 𝐴) = (𝐴 × { 0 })}    &   ⊢ 𝐹 = (𝑥 ∈ 𝐾 ↦ (𝑥 ↾ 𝐵))    &   ⊢ 𝐶 = (𝑅 ↑s 𝐴)    &   ⊢ 𝐷 = (𝑅 ↑s 𝐵)    &   ⊢ 𝐿 = (𝐸 ↾s 𝐾)    ⇒   ⊢ ((𝑅 ∈ LMod ∧ (𝐴 ∪ 𝐵) ∈ 𝑉 ∧ (𝐴 ∩ 𝐵) = ∅) → 𝐹 ∈ (𝐿 LMIso 𝐷))
Theorem	filnm 39188	Finite left modules are Noetherian. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝐵 = (Base‘𝑊)    ⇒   ⊢ ((𝑊 ∈ LMod ∧ 𝐵 ∈ Fin) → 𝑊 ∈ LNoeM)
Theorem	pwslnmlem0 39189	Zeroeth powers are Noetherian. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑌 = (𝑊 ↑s ∅)    ⇒   ⊢ (𝑊 ∈ LMod → 𝑌 ∈ LNoeM)
Theorem	pwslnmlem1 39190*	First powers are Noetherian. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑌 = (𝑊 ↑s {𝑖})    ⇒   ⊢ (𝑊 ∈ LNoeM → 𝑌 ∈ LNoeM)
Theorem	pwslnmlem2 39191	A sum of powers is Noetherian. (Contributed by Stefan O'Rear, 25-Jan-2015.)
⊢ 𝐴 ∈ V    &   ⊢ 𝐵 ∈ V    &   ⊢ 𝑋 = (𝑊 ↑s 𝐴)    &   ⊢ 𝑌 = (𝑊 ↑s 𝐵)    &   ⊢ 𝑍 = (𝑊 ↑s (𝐴 ∪ 𝐵))    &   ⊢ (𝜑 → 𝑊 ∈ LMod)    &   ⊢ (𝜑 → (𝐴 ∩ 𝐵) = ∅)    &   ⊢ (𝜑 → 𝑋 ∈ LNoeM)    &   ⊢ (𝜑 → 𝑌 ∈ LNoeM)    ⇒   ⊢ (𝜑 → 𝑍 ∈ LNoeM)
Theorem	pwslnm 39192	Finite powers of Noetherian modules are Noetherian. (Contributed by Stefan O'Rear, 24-Jan-2015.)
⊢ 𝑌 = (𝑊 ↑s 𝐼)    ⇒   ⊢ ((𝑊 ∈ LNoeM ∧ 𝐼 ∈ Fin) → 𝑌 ∈ LNoeM)
20.27.40  Every set admits a group structure iff choice
Theorem	unxpwdom3 39193*	Weaker version of unxpwdom 8902 where a function is required only to be cancellative, not an injection. 𝐷 and 𝐵 are to be thought of as "large" "horizonal" sets, the others as "small". Because the operator is row-wise injective, but the whole row cannot inject into 𝐴, each row must hit an element of 𝐵; by column injectivity, each row can be identified in at least one way by the 𝐵 element that it hits and the column in which it is hit. (Contributed by Stefan O'Rear, 8-Jul-2015.) MOVABLE
⊢ (𝜑 → 𝐴 ∈ 𝑉)    &   ⊢ (𝜑 → 𝐵 ∈ 𝑊)    &   ⊢ (𝜑 → 𝐷 ∈ 𝑋)    &   ⊢ ((𝜑 ∧ 𝑎 ∈ 𝐶 ∧ 𝑏 ∈ 𝐷) → (𝑎 + 𝑏) ∈ (𝐴 ∪ 𝐵))    &   ⊢ (((𝜑 ∧ 𝑎 ∈ 𝐶) ∧ (𝑏 ∈ 𝐷 ∧ 𝑐 ∈ 𝐷)) → ((𝑎 + 𝑏) = (𝑎 + 𝑐) ↔ 𝑏 = 𝑐))    &   ⊢ (((𝜑 ∧ 𝑑 ∈ 𝐷) ∧ (𝑎 ∈ 𝐶 ∧ 𝑐 ∈ 𝐶)) → ((𝑐 + 𝑑) = (𝑎 + 𝑑) ↔ 𝑐 = 𝑎))    &   ⊢ (𝜑 → ¬ 𝐷 ≼ 𝐴)    ⇒   ⊢ (𝜑 → 𝐶 ≼* (𝐷 × 𝐵))
Theorem	pwfi2f1o 39194*	The pw2f1o 8472 bijection relates finitely supported indicator functions on a two-element set to finite subsets. MOVABLE (Contributed by Stefan O'Rear, 10-Jul-2015.) (Revised by AV, 14-Jun-2020.)
⊢ 𝑆 = {𝑦 ∈ (2o ↑𝑚 𝐴) ∣ 𝑦 finSupp ∅}    &   ⊢ 𝐹 = (𝑥 ∈ 𝑆 ↦ (◡𝑥 “ {1o}))    ⇒   ⊢ (𝐴 ∈ 𝑉 → 𝐹:𝑆–1-1-onto→(𝒫 𝐴 ∩ Fin))
Theorem	pwfi2en 39195*	Finitely supported indicator functions are equinumerous to finite subsets. MOVABLE (Contributed by Stefan O'Rear, 10-Jul-2015.) (Revised by AV, 14-Jun-2020.)
⊢ 𝑆 = {𝑦 ∈ (2o ↑𝑚 𝐴) ∣ 𝑦 finSupp ∅}    ⇒   ⊢ (𝐴 ∈ 𝑉 → 𝑆 ≈ (𝒫 𝐴 ∩ Fin))
Theorem	frlmpwfi 39196	Formal linear combinations over Z/2Z are equivalent to finite subsets. MOVABLE (Contributed by Stefan O'Rear, 10-Jul-2015.) (Proof shortened by AV, 14-Jun-2020.)
⊢ 𝑅 = (ℤ/nℤ‘2)    &   ⊢ 𝑌 = (𝑅 freeLMod 𝐼)    &   ⊢ 𝐵 = (Base‘𝑌)    ⇒   ⊢ (𝐼 ∈ 𝑉 → 𝐵 ≈ (𝒫 𝐼 ∩ Fin))
Theorem	gicabl 39197	Being Abelian is a group invariant. MOVABLE (Contributed by Stefan O'Rear, 8-Jul-2015.)
⊢ (𝐺 ≃𝑔 𝐻 → (𝐺 ∈ Abel ↔ 𝐻 ∈ Abel))
Theorem	imasgim 39198	A relabeling of the elements of a group induces an isomorphism to the relabeled group. MOVABLE (Contributed by Stefan O'Rear, 8-Jul-2015.) (Revised by Mario Carneiro, 11-Aug-2015.)
⊢ (𝜑 → 𝑈 = (𝐹 “s 𝑅))    &   ⊢ (𝜑 → 𝑉 = (Base‘𝑅))    &   ⊢ (𝜑 → 𝐹:𝑉–1-1-onto→𝐵)    &   ⊢ (𝜑 → 𝑅 ∈ Grp)    ⇒   ⊢ (𝜑 → 𝐹 ∈ (𝑅 GrpIso 𝑈))
Theorem	isnumbasgrplem1 39199	A set which is equipollent to the base set of a definable Abelian group is the base set of some (relabeled) Abelian group. (Contributed by Stefan O'Rear, 8-Jul-2015.)
⊢ 𝐵 = (Base‘𝑅)    ⇒   ⊢ ((𝑅 ∈ Abel ∧ 𝐶 ≈ 𝐵) → 𝐶 ∈ (Base “ Abel))
Theorem	harn0 39200	The Hartogs number of a set is never zero. MOVABLE (Contributed by Stefan O'Rear, 9-Jul-2015.)
⊢ (𝑆 ∈ 𝑉 → (har‘𝑆) ≠ ∅)