P. 476 of Theorem List - Metamath Proof Explorer

Theorem List for Metamath Proof Explorer - 47501-47600   *Has distinct variable group(s)

Type	Label	Description
Statement
Theorem	setsnidel 47501	The injected slot is an element of the structure with replacement. (Contributed by AV, 10-Nov-2021.)
⊢ (𝜑 → 𝑆 ∈ 𝑉)    &   ⊢ (𝜑 → 𝐵 ∈ 𝑊)    &   ⊢ 𝑅 = (𝑆 sSet ⟨𝐴, 𝐵⟩)    &   ⊢ (𝜑 → 𝐶 ∈ 𝑋)    &   ⊢ (𝜑 → 𝐷 ∈ 𝑌)    &   ⊢ (𝜑 → ⟨𝐶, 𝐷⟩ ∈ 𝑆)    &   ⊢ (𝜑 → 𝐴 ≠ 𝐶)    ⇒   ⊢ (𝜑 → ⟨𝐶, 𝐷⟩ ∈ 𝑅)
Theorem	setsv 47502	The value of the structure replacement function is a set. (Contributed by AV, 10-Nov-2021.)
⊢ ((𝑆 ∈ 𝑉 ∧ 𝐵 ∈ 𝑊) → (𝑆 sSet ⟨𝐴, 𝐵⟩) ∈ V)
21.48.7  Preimages of function values According to Wikipedia ("Image (mathematics)", 17-Mar-2024, https://en.wikipedia.org/wiki/ImageSupport_(mathematics)): "... evaluating a given function 𝑓 at each element of a given subset 𝐴 of its domain produces a set, called the "image of 𝐴 under (or through) 𝑓". Similarly, the inverse image (or preimage) of a given subset 𝐵 of the codomain of 𝑓 is the set of all elements of the domain that map to the members of 𝐵." The preimage of a set 𝐵 under a function 𝑓 is often denoted as "f^-1 (B)", but in set.mm, the idiom (◡𝑓 “ 𝐵) is used. As a special case, the idiom for the preimage of a function value at 𝑋 under a function 𝐹 is (◡𝐹 “ {(𝐹‘𝑋)}) (according to Wikipedia, the preimage of a singleton is also called a "fiber"). We use the label fragment "preima" (as in mptpreima 6190) for theorems about preimages (sometimes, also "imacnv" is used as in fvimacnvi 6991), and "preimafv" (as in preimafvn0 47504) for theorems about preimages of a function value. In this section, 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})} will be the set of all preimages of function values of a function 𝐹, that means 𝑆 ∈ 𝑃 is a preimage of a function value (see, for example, elsetpreimafv 47509): 𝑆 = (◡𝐹 “ {(𝐹‘𝑥)}). With the help of such a set, it is shown that every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective and an injective function (see fundcmpsurinj 47533) by constructing a surjective function 𝑔:𝐴–onto→𝑃 and an injective function ℎ:𝑃–1-1→𝐵 so that 𝐹 = (ℎ ∘ 𝑔) ( see fundcmpsurinjpreimafv 47532). See also Wikipedia ("Surjective function", 17-Mar-2024, https://en.wikipedia.org/wiki/Surjective_function 47532 (section "Composition and decomposition"). This is different from the decomposition of 𝐹 into the surjective function 𝑔:𝐴–onto→(𝐹 “ 𝐴) (with (𝑔‘𝑥) = (𝐹‘𝑥) for 𝑥 ∈ 𝐴) and the injective function ℎ = ( I ↾ (𝐹 “ 𝐴)), ( see fundcmpsurinjimaid 47535), see also Wikipedia ("Bijection, injection and surjection", 17-Mar-2024, https://en.wikipedia.org/wiki/Bijection,_injection_and_surjection 47535 (section "Properties"). Finally, it is shown that every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective, a bijective and an injective function (see fundcmpsurbijinj 47534), by showing that there is a bijection between the set of all preimages of values of a function and the range of the function (see imasetpreimafvbij 47530). From this, both variants of decompositions of a function into a surjective and an injective function can be derived: Let 𝐹 = ((𝐼 ∘ 𝐵) ∘ 𝑆) be a decomposition of a function into a surjective, a bijective and an injective function, then 𝐹 = (𝐽 ∘ 𝑆) with 𝐽 = (𝐼 ∘ 𝐵) (an injective function) is a decomposition into a surjective and an injective function corresponding to fundcmpsurinj 47533, and 𝐹 = (𝐼 ∘ 𝑂) with 𝑂 = (𝐵 ∘ 𝑆) (a surjective function) is a decomposition into a surjective and an injective function corresponding to fundcmpsurinjimaid 47535.
Theorem	preimafvsnel 47503	The preimage of a function value at 𝑋 contains 𝑋. (Contributed by AV, 7-Mar-2024.)
⊢ ((𝐹 Fn 𝐴 ∧ 𝑋 ∈ 𝐴) → 𝑋 ∈ (◡𝐹 “ {(𝐹‘𝑋)}))
Theorem	preimafvn0 47504	The preimage of a function value is not empty. (Contributed by AV, 7-Mar-2024.)
⊢ ((𝐹 Fn 𝐴 ∧ 𝑋 ∈ 𝐴) → (◡𝐹 “ {(𝐹‘𝑋)}) ≠ ∅)
Theorem	uniimafveqt 47505*	The union of the image of a subset 𝑆 of the domain of a function with elements having the same function value is the function value at one of the elements of 𝑆. (Contributed by AV, 5-Mar-2024.)
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝑆 ⊆ 𝐴 ∧ 𝑋 ∈ 𝑆) → (∀𝑥 ∈ 𝑆 (𝐹‘𝑥) = (𝐹‘𝑋) → ∪ (𝐹 “ 𝑆) = (𝐹‘𝑋)))
Theorem	uniimaprimaeqfv 47506	The union of the image of the preimage of a function value is the function value. (Contributed by AV, 12-Mar-2024.)
⊢ ((𝐹 Fn 𝐴 ∧ 𝑋 ∈ 𝐴) → ∪ (𝐹 “ (◡𝐹 “ {(𝐹‘𝑋)})) = (𝐹‘𝑋))
Theorem	setpreimafvex 47507*	The class 𝑃 of all preimages of function values is a set. (Contributed by AV, 10-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ (𝐴 ∈ 𝑉 → 𝑃 ∈ V)
Theorem	elsetpreimafvb 47508*	The characterization of an element of the class 𝑃 of all preimages of function values. (Contributed by AV, 10-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ (𝑆 ∈ 𝑉 → (𝑆 ∈ 𝑃 ↔ ∃𝑥 ∈ 𝐴 𝑆 = (◡𝐹 “ {(𝐹‘𝑥)})))
Theorem	elsetpreimafv 47509*	An element of the class 𝑃 of all preimages of function values. (Contributed by AV, 8-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ (𝑆 ∈ 𝑃 → ∃𝑥 ∈ 𝐴 𝑆 = (◡𝐹 “ {(𝐹‘𝑥)}))
Theorem	elsetpreimafvssdm 47510*	An element of the class 𝑃 of all preimages of function values is a subset of the domain of the function. (Contributed by AV, 8-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃) → 𝑆 ⊆ 𝐴)
Theorem	fvelsetpreimafv 47511*	There is an element in a preimage 𝑆 of function values so that 𝑆 is the preimage of the function value at this element. (Contributed by AV, 8-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃) → ∃𝑥 ∈ 𝑆 𝑆 = (◡𝐹 “ {(𝐹‘𝑥)}))
Theorem	preimafvelsetpreimafv 47512*	The preimage of a function value is an element of the class 𝑃 of all preimages of function values. (Contributed by AV, 10-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝐴 ∈ 𝑉 ∧ 𝑋 ∈ 𝐴) → (◡𝐹 “ {(𝐹‘𝑋)}) ∈ 𝑃)
Theorem	preimafvsspwdm 47513*	The class 𝑃 of all preimages of function values is a subset of the power set of the domain of the function. (Contributed by AV, 5-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ (𝐹 Fn 𝐴 → 𝑃 ⊆ 𝒫 𝐴)
Theorem	0nelsetpreimafv 47514*	The empty set is not an element of the class 𝑃 of all preimages of function values. (Contributed by AV, 6-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ (𝐹 Fn 𝐴 → ∅ ∉ 𝑃)
Theorem	elsetpreimafvbi 47515*	An element of the preimage of a function value is an element of the domain of the function with the same value as another element of the preimage. (Contributed by AV, 9-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃 ∧ 𝑋 ∈ 𝑆) → (𝑌 ∈ 𝑆 ↔ (𝑌 ∈ 𝐴 ∧ (𝐹‘𝑌) = (𝐹‘𝑋))))
Theorem	elsetpreimafveqfv 47516*	The elements of the preimage of a function value have the same function values. (Contributed by AV, 5-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ (𝑆 ∈ 𝑃 ∧ 𝑋 ∈ 𝑆 ∧ 𝑌 ∈ 𝑆)) → (𝐹‘𝑋) = (𝐹‘𝑌))
Theorem	eqfvelsetpreimafv 47517*	If an element of the domain of the function has the same function value as an element of the preimage of a function value, then it is an element of the same preimage. (Contributed by AV, 9-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃 ∧ 𝑋 ∈ 𝑆) → ((𝑌 ∈ 𝐴 ∧ (𝐹‘𝑌) = (𝐹‘𝑋)) → 𝑌 ∈ 𝑆))
Theorem	elsetpreimafvrab 47518*	An element of the preimage of a function value expressed as a restricted class abstraction. (Contributed by AV, 9-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃 ∧ 𝑋 ∈ 𝑆) → 𝑆 = {𝑥 ∈ 𝐴 ∣ (𝐹‘𝑥) = (𝐹‘𝑋)})
Theorem	imaelsetpreimafv 47519*	The image of an element of the preimage of a function value is the singleton consisting of the function value at one of its elements. (Contributed by AV, 5-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃 ∧ 𝑋 ∈ 𝑆) → (𝐹 “ 𝑆) = {(𝐹‘𝑋)})
Theorem	uniimaelsetpreimafv 47520*	The union of the image of an element of the preimage of a function value is an element of the range of the function. (Contributed by AV, 5-Mar-2024.) (Revised by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑆 ∈ 𝑃) → ∪ (𝐹 “ 𝑆) ∈ ran 𝐹)
Theorem	elsetpreimafveq 47521*	If two preimages of function values contain elements with identical function values, then both preimages are equal. (Contributed by AV, 8-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ (𝑆 ∈ 𝑃 ∧ 𝑅 ∈ 𝑃) ∧ (𝑋 ∈ 𝑆 ∧ 𝑌 ∈ 𝑅)) → ((𝐹‘𝑋) = (𝐹‘𝑌) → 𝑆 = 𝑅))
Theorem	fundcmpsurinjlem1 47522*	Lemma 1 for fundcmpsurinj 47533. (Contributed by AV, 4-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐺 = (𝑥 ∈ 𝐴 ↦ (◡𝐹 “ {(𝐹‘𝑥)}))    ⇒   ⊢ ran 𝐺 = 𝑃
Theorem	fundcmpsurinjlem2 47523*	Lemma 2 for fundcmpsurinj 47533. (Contributed by AV, 4-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐺 = (𝑥 ∈ 𝐴 ↦ (◡𝐹 “ {(𝐹‘𝑥)}))    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝐴 ∈ 𝑉) → 𝐺:𝐴–onto→𝑃)
Theorem	fundcmpsurinjlem3 47524*	Lemma 3 for fundcmpsurinj 47533. (Contributed by AV, 3-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ ((Fun 𝐹 ∧ 𝑋 ∈ 𝑃) → (𝐻‘𝑋) = ∪ (𝐹 “ 𝑋))
Theorem	imasetpreimafvbijlemf 47525*	Lemma for imasetpreimafvbij 47530: the mapping 𝐻 is a function into the range of function 𝐹. (Contributed by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ (𝐹 Fn 𝐴 → 𝐻:𝑃⟶(𝐹 “ 𝐴))
Theorem	imasetpreimafvbijlemfv 47526*	Lemma for imasetpreimafvbij 47530: the value of the mapping 𝐻 at a preimage of a value of function 𝐹. (Contributed by AV, 5-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑌 ∈ 𝑃 ∧ 𝑋 ∈ 𝑌) → (𝐻‘𝑌) = (𝐹‘𝑋))
Theorem	imasetpreimafvbijlemfv1 47527*	Lemma for imasetpreimafvbij 47530: for a preimage of a value of function 𝐹 there is an element of the preimage so that the value of the mapping 𝐻 at this preimage is the function value at this element. (Contributed by AV, 5-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝑋 ∈ 𝑃) → ∃𝑦 ∈ 𝑋 (𝐻‘𝑋) = (𝐹‘𝑦))
Theorem	imasetpreimafvbijlemf1 47528*	Lemma for imasetpreimafvbij 47530: the mapping 𝐻 is an injective function into the range of function 𝐹. (Contributed by AV, 9-Mar-2024.) (Revised by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ (𝐹 Fn 𝐴 → 𝐻:𝑃–1-1→(𝐹 “ 𝐴))
Theorem	imasetpreimafvbijlemfo 47529*	Lemma for imasetpreimafvbij 47530: the mapping 𝐻 is a function onto the range of function 𝐹. (Contributed by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝐴 ∈ 𝑉) → 𝐻:𝑃–onto→(𝐹 “ 𝐴))
Theorem	imasetpreimafvbij 47530*	The mapping 𝐻 is a bijective function between the set 𝑃 of all preimages of values of function 𝐹 and the range of 𝐹. (Contributed by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    &   ⊢ 𝐻 = (𝑝 ∈ 𝑃 ↦ ∪ (𝐹 “ 𝑝))    ⇒   ⊢ ((𝐹 Fn 𝐴 ∧ 𝐴 ∈ 𝑉) → 𝐻:𝑃–1-1-onto→(𝐹 “ 𝐴))
Theorem	fundcmpsurbijinjpreimafv 47531*	Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective function onto 𝑃, a bijective function from 𝑃 and an injective function into the codomain of 𝐹. (Contributed by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐴 ∈ 𝑉) → ∃𝑔∃ℎ∃𝑖((𝑔:𝐴–onto→𝑃 ∧ ℎ:𝑃–1-1-onto→(𝐹 “ 𝐴) ∧ 𝑖:(𝐹 “ 𝐴)–1-1→𝐵) ∧ 𝐹 = ((𝑖 ∘ ℎ) ∘ 𝑔)))
Theorem	fundcmpsurinjpreimafv 47532*	Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective function onto 𝑃 and an injective function from 𝑃. (Contributed by AV, 12-Mar-2024.) (Proof shortened by AV, 22-Mar-2024.)
⊢ 𝑃 = {𝑧 ∣ ∃𝑥 ∈ 𝐴 𝑧 = (◡𝐹 “ {(𝐹‘𝑥)})}    ⇒   ⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐴 ∈ 𝑉) → ∃𝑔∃ℎ(𝑔:𝐴–onto→𝑃 ∧ ℎ:𝑃–1-1→𝐵 ∧ 𝐹 = (ℎ ∘ 𝑔)))
Theorem	fundcmpsurinj 47533*	Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective and an injective function. (Contributed by AV, 13-Mar-2024.)
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐴 ∈ 𝑉) → ∃𝑔∃ℎ∃𝑝(𝑔:𝐴–onto→𝑝 ∧ ℎ:𝑝–1-1→𝐵 ∧ 𝐹 = (ℎ ∘ 𝑔)))
Theorem	fundcmpsurbijinj 47534*	Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective, a bijective and an injective function. (Contributed by AV, 23-Mar-2024.)
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐴 ∈ 𝑉) → ∃𝑔∃ℎ∃𝑖∃𝑝∃𝑞((𝑔:𝐴–onto→𝑝 ∧ ℎ:𝑝–1-1-onto→𝑞 ∧ 𝑖:𝑞–1-1→𝐵) ∧ 𝐹 = ((𝑖 ∘ ℎ) ∘ 𝑔)))
Theorem	fundcmpsurinjimaid 47535*	Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective function onto the image (𝐹 “ 𝐴) of the domain of 𝐹 and an injective function from the image (𝐹 “ 𝐴). (Contributed by AV, 17-Mar-2024.)
⊢ 𝐼 = (𝐹 “ 𝐴)    &   ⊢ 𝐺 = (𝑥 ∈ 𝐴 ↦ (𝐹‘𝑥))    &   ⊢ 𝐻 = ( I ↾ 𝐼)    ⇒   ⊢ (𝐹:𝐴⟶𝐵 → (𝐺:𝐴–onto→𝐼 ∧ 𝐻:𝐼–1-1→𝐵 ∧ 𝐹 = (𝐻 ∘ 𝐺)))
Theorem	fundcmpsurinjALT 47536*	Alternate proof of fundcmpsurinj 47533, based on fundcmpsurinjimaid 47535: Every function 𝐹:𝐴⟶𝐵 can be decomposed into a surjective and an injective function. (Proof modification is discouraged.) (New usage is discouraged.) (Contributed by AV, 13-Mar-2024.)
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐴 ∈ 𝑉) → ∃𝑔∃ℎ∃𝑝(𝑔:𝐴–onto→𝑝 ∧ ℎ:𝑝–1-1→𝐵 ∧ 𝐹 = (ℎ ∘ 𝑔)))
21.48.8  Partitions of real intervals Based on the theorems of the fourierdlem* series of GS's mathbox.
Syntax	ciccp 47537	Extend class notation with the partitions of a closed interval of extended reals.
class RePart
Definition	df-iccp 47538*	Define partitions of a closed interval of extended reals. Such partitions are finite increasing sequences of extended reals. (Contributed by AV, 8-Jul-2020.)
⊢ RePart = (𝑚 ∈ ℕ ↦ {𝑝 ∈ (ℝ* ↑m (0...𝑚)) ∣ ∀𝑖 ∈ (0..^𝑚)(𝑝‘𝑖) < (𝑝‘(𝑖 + 1))})
Theorem	iccpval 47539*	Partition consisting of a fixed number 𝑀 of parts. (Contributed by AV, 9-Jul-2020.)
⊢ (𝑀 ∈ ℕ → (RePart‘𝑀) = {𝑝 ∈ (ℝ* ↑m (0...𝑀)) ∣ ∀𝑖 ∈ (0..^𝑀)(𝑝‘𝑖) < (𝑝‘(𝑖 + 1))})
Theorem	iccpart 47540*	A special partition. Corresponds to fourierdlem2 46231 in GS's mathbox. (Contributed by AV, 9-Jul-2020.)
⊢ (𝑀 ∈ ℕ → (𝑃 ∈ (RePart‘𝑀) ↔ (𝑃 ∈ (ℝ* ↑m (0...𝑀)) ∧ ∀𝑖 ∈ (0..^𝑀)(𝑃‘𝑖) < (𝑃‘(𝑖 + 1)))))
Theorem	iccpartimp 47541	Implications for a class being a partition. (Contributed by AV, 11-Jul-2020.)
⊢ ((𝑀 ∈ ℕ ∧ 𝑃 ∈ (RePart‘𝑀) ∧ 𝐼 ∈ (0..^𝑀)) → (𝑃 ∈ (ℝ* ↑m (0...𝑀)) ∧ (𝑃‘𝐼) < (𝑃‘(𝐼 + 1))))
Theorem	iccpartres 47542	The restriction of a partition is a partition. (Contributed by AV, 16-Jul-2020.)
⊢ ((𝑀 ∈ ℕ ∧ 𝑃 ∈ (RePart‘(𝑀 + 1))) → (𝑃 ↾ (0...𝑀)) ∈ (RePart‘𝑀))
Theorem	iccpartxr 47543	If there is a partition, then all intermediate points and bounds are extended real numbers. (Contributed by AV, 11-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    &   ⊢ (𝜑 → 𝐼 ∈ (0...𝑀))    ⇒   ⊢ (𝜑 → (𝑃‘𝐼) ∈ ℝ*)
Theorem	iccpartgtprec 47544	If there is a partition, then all intermediate points and the upper bound are strictly greater than the preceeding intermediate points or lower bound. (Contributed by AV, 11-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    &   ⊢ (𝜑 → 𝐼 ∈ (1...𝑀))    ⇒   ⊢ (𝜑 → (𝑃‘(𝐼 − 1)) < (𝑃‘𝐼))
Theorem	iccpartipre 47545	If there is a partition, then all intermediate points are real numbers. (Contributed by AV, 11-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    &   ⊢ (𝜑 → 𝐼 ∈ (1..^𝑀))    ⇒   ⊢ (𝜑 → (𝑃‘𝐼) ∈ ℝ)
Theorem	iccpartiltu 47546*	If there is a partition, then all intermediate points are strictly less than the upper bound. (Contributed by AV, 12-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (1..^𝑀)(𝑃‘𝑖) < (𝑃‘𝑀))
Theorem	iccpartigtl 47547*	If there is a partition, then all intermediate points are strictly greater than the lower bound. (Contributed by AV, 12-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (1..^𝑀)(𝑃‘0) < (𝑃‘𝑖))
Theorem	iccpartlt 47548	If there is a partition, then the lower bound is strictly less than the upper bound. Corresponds to fourierdlem11 46240 in GS's mathbox. (Contributed by AV, 12-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → (𝑃‘0) < (𝑃‘𝑀))
Theorem	iccpartltu 47549*	If there is a partition, then all intermediate points and the lower bound are strictly less than the upper bound. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (0..^𝑀)(𝑃‘𝑖) < (𝑃‘𝑀))
Theorem	iccpartgtl 47550*	If there is a partition, then all intermediate points and the upper bound are strictly greater than the lower bound. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (1...𝑀)(𝑃‘0) < (𝑃‘𝑖))
Theorem	iccpartgt 47551*	If there is a partition, then all intermediate points and the bounds are strictly ordered. (Contributed by AV, 18-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (0...𝑀)∀𝑗 ∈ (0...𝑀)(𝑖 < 𝑗 → (𝑃‘𝑖) < (𝑃‘𝑗)))
Theorem	iccpartleu 47552*	If there is a partition, then all intermediate points and the lower and the upper bound are less than or equal to the upper bound. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (0...𝑀)(𝑃‘𝑖) ≤ (𝑃‘𝑀))
Theorem	iccpartgel 47553*	If there is a partition, then all intermediate points and the upper and the lower bound are greater than or equal to the lower bound. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ∀𝑖 ∈ (0...𝑀)(𝑃‘0) ≤ (𝑃‘𝑖))
Theorem	iccpartrn 47554	If there is a partition, then all intermediate points and bounds are contained in a closed interval of extended reals. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ran 𝑃 ⊆ ((𝑃‘0)[,](𝑃‘𝑀)))
Theorem	iccpartf 47555	The range of the partition is between its starting point and its ending point. Corresponds to fourierdlem15 46244 in GS's mathbox. (Contributed by Glauco Siliprandi, 11-Dec-2019.) (Revised by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → 𝑃:(0...𝑀)⟶((𝑃‘0)[,](𝑃‘𝑀)))
Theorem	iccpartel 47556	If there is a partition, then all intermediate points and bounds are contained in a closed interval of extended reals. (Contributed by AV, 14-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ ((𝜑 ∧ 𝐼 ∈ (0...𝑀)) → (𝑃‘𝐼) ∈ ((𝑃‘0)[,](𝑃‘𝑀)))
Theorem	iccelpart 47557*	An element of any partitioned half-open interval of extended reals is an element of a part of this partition. (Contributed by AV, 18-Jul-2020.)
⊢ (𝑀 ∈ ℕ → ∀𝑝 ∈ (RePart‘𝑀)(𝑋 ∈ ((𝑝‘0)[,)(𝑝‘𝑀)) → ∃𝑖 ∈ (0..^𝑀)𝑋 ∈ ((𝑝‘𝑖)[,)(𝑝‘(𝑖 + 1)))))
Theorem	iccpartiun 47558*	A half-open interval of extended reals is the union of the parts of its partition. (Contributed by AV, 18-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ((𝑃‘0)[,)(𝑃‘𝑀)) = ∪ 𝑖 ∈ (0..^𝑀)((𝑃‘𝑖)[,)(𝑃‘(𝑖 + 1))))
Theorem	icceuelpartlem 47559	Lemma for icceuelpart 47560. (Contributed by AV, 19-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → ((𝐼 ∈ (0..^𝑀) ∧ 𝐽 ∈ (0..^𝑀)) → (𝐼 < 𝐽 → (𝑃‘(𝐼 + 1)) ≤ (𝑃‘𝐽))))
Theorem	icceuelpart 47560*	An element of a partitioned half-open interval of extended reals is an element of exactly one part of the partition. (Contributed by AV, 19-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ ((𝜑 ∧ 𝑋 ∈ ((𝑃‘0)[,)(𝑃‘𝑀))) → ∃!𝑖 ∈ (0..^𝑀)𝑋 ∈ ((𝑃‘𝑖)[,)(𝑃‘(𝑖 + 1))))
Theorem	iccpartdisj 47561*	The segments of a partitioned half-open interval of extended reals are a disjoint collection. (Contributed by AV, 19-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    ⇒   ⊢ (𝜑 → Disj 𝑖 ∈ (0..^𝑀)((𝑃‘𝑖)[,)(𝑃‘(𝑖 + 1))))
Theorem	iccpartnel 47562	A point of a partition is not an element of any open interval determined by the partition. Corresponds to fourierdlem12 46241 in GS's mathbox. (Contributed by Glauco Siliprandi, 11-Dec-2019.) (Revised by AV, 8-Jul-2020.)
⊢ (𝜑 → 𝑀 ∈ ℕ)    &   ⊢ (𝜑 → 𝑃 ∈ (RePart‘𝑀))    &   ⊢ (𝜑 → 𝑋 ∈ ran 𝑃)    ⇒   ⊢ ((𝜑 ∧ 𝐼 ∈ (0..^𝑀)) → ¬ 𝑋 ∈ ((𝑃‘𝐼)(,)(𝑃‘(𝐼 + 1))))
21.48.9  Shifting functions with an integer range domain
Theorem	fargshiftfv 47563*	If a class is a function, then the values of the "shifted function" correspond to the function values of the class. (Contributed by Alexander van der Vekens, 23-Nov-2017.)
⊢ 𝐺 = (𝑥 ∈ (0..^(♯‘𝐹)) ↦ (𝐹‘(𝑥 + 1)))    ⇒   ⊢ ((𝑁 ∈ ℕ0 ∧ 𝐹:(1...𝑁)⟶dom 𝐸) → (𝑋 ∈ (0..^𝑁) → (𝐺‘𝑋) = (𝐹‘(𝑋 + 1))))
Theorem	fargshiftf 47564*	If a class is a function, then also its "shifted function" is a function. (Contributed by Alexander van der Vekens, 23-Nov-2017.)
⊢ 𝐺 = (𝑥 ∈ (0..^(♯‘𝐹)) ↦ (𝐹‘(𝑥 + 1)))    ⇒   ⊢ ((𝑁 ∈ ℕ0 ∧ 𝐹:(1...𝑁)⟶dom 𝐸) → 𝐺:(0..^(♯‘𝐹))⟶dom 𝐸)
Theorem	fargshiftf1 47565*	If a function is 1-1, then also the shifted function is 1-1. (Contributed by Alexander van der Vekens, 23-Nov-2017.)
⊢ 𝐺 = (𝑥 ∈ (0..^(♯‘𝐹)) ↦ (𝐹‘(𝑥 + 1)))    ⇒   ⊢ ((𝑁 ∈ ℕ0 ∧ 𝐹:(1...𝑁)–1-1→dom 𝐸) → 𝐺:(0..^(♯‘𝐹))–1-1→dom 𝐸)
Theorem	fargshiftfo 47566*	If a function is onto, then also the shifted function is onto. (Contributed by Alexander van der Vekens, 24-Nov-2017.)
⊢ 𝐺 = (𝑥 ∈ (0..^(♯‘𝐹)) ↦ (𝐹‘(𝑥 + 1)))    ⇒   ⊢ ((𝑁 ∈ ℕ0 ∧ 𝐹:(1...𝑁)–onto→dom 𝐸) → 𝐺:(0..^(♯‘𝐹))–onto→dom 𝐸)
Theorem	fargshiftfva 47567*	The values of a shifted function correspond to the value of the original function. (Contributed by Alexander van der Vekens, 24-Nov-2017.)
⊢ 𝐺 = (𝑥 ∈ (0..^(♯‘𝐹)) ↦ (𝐹‘(𝑥 + 1)))    ⇒   ⊢ ((𝑁 ∈ ℕ0 ∧ 𝐹:(1...𝑁)⟶dom 𝐸) → (∀𝑘 ∈ (1...𝑁)(𝐸‘(𝐹‘𝑘)) = ⦋𝑘 / 𝑥⦌𝑃 → ∀𝑙 ∈ (0..^𝑁)(𝐸‘(𝐺‘𝑙)) = ⦋(𝑙 + 1) / 𝑥⦌𝑃))
21.48.10  Words over a set (extension)
21.48.10.1  Last symbol of a word - extension
Theorem	lswn0 47568	The last symbol of a nonempty word exists. The empty set must be excluded as symbol, because otherwise, it cannot be distinguished between valid cases (∅ is the last symbol) and invalid cases (∅ means that no last symbol exists). This is because of the special definition of a function in set.mm. (Contributed by Alexander van der Vekens, 18-Mar-2018.)
⊢ ((𝑊 ∈ Word 𝑉 ∧ ∅ ∉ 𝑉 ∧ (♯‘𝑊) ≠ 0) → (lastS‘𝑊) ≠ ∅)
21.48.11  Unordered pairs
21.48.11.1  Interchangeable setvar variables
Syntax	wich 47569	Extend wff notation to include the property of a wff 𝜑 that the setvar variables 𝑥 and 𝑦 are interchangeable. Read this notation as "𝑥 and 𝑦 are interchangeable in wff 𝜑".
wff [𝑥⇄𝑦]𝜑
Definition	df-ich 47570*	Define the property of a wff 𝜑 that the setvar variables 𝑥 and 𝑦 are interchangeable. For an alternate definition using implicit substitution and a temporary setvar variable see ichcircshi 47578. Another, equivalent definition using two temporary setvar variables is provided in dfich2 47582. (Contributed by AV, 29-Jul-2023.)
⊢ ([𝑥⇄𝑦]𝜑 ↔ ∀𝑥∀𝑦([𝑥 / 𝑎][𝑦 / 𝑥][𝑎 / 𝑦]𝜑 ↔ 𝜑))
Theorem	nfich1 47571	The first interchangeable setvar variable is not free. (Contributed by AV, 21-Aug-2023.)
⊢ Ⅎ𝑥[𝑥⇄𝑦]𝜑
Theorem	nfich2 47572	The second interchangeable setvar variable is not free. (Contributed by AV, 21-Aug-2023.)
⊢ Ⅎ𝑦[𝑥⇄𝑦]𝜑
Theorem	ichv 47573*	Setvar variables are interchangeable in a wff they do not appear in. (Contributed by SN, 23-Nov-2023.)
⊢ [𝑥⇄𝑦]𝜑
Theorem	ichf 47574	Setvar variables are interchangeable in a wff they are not free in. (Contributed by SN, 23-Nov-2023.)
⊢ Ⅎ𝑥𝜑    &   ⊢ Ⅎ𝑦𝜑    ⇒   ⊢ [𝑥⇄𝑦]𝜑
Theorem	ichid 47575	A setvar variable is always interchangeable with itself. (Contributed by AV, 29-Jul-2023.)
⊢ [𝑥⇄𝑥]𝜑
Theorem	icht 47576	A theorem is interchangeable. (Contributed by SN, 4-May-2024.)
⊢ 𝜑    ⇒   ⊢ [𝑥⇄𝑦]𝜑
Theorem	ichbidv 47577*	Formula building rule for interchangeability (deduction). (Contributed by SN, 4-May-2024.)
⊢ (𝜑 → (𝜓 ↔ 𝜒))    ⇒   ⊢ (𝜑 → ([𝑥⇄𝑦]𝜓 ↔ [𝑥⇄𝑦]𝜒))
Theorem	ichcircshi 47578*	The setvar variables are interchangeable if they can be circularily shifted using a third setvar variable, using implicit substitution. (Contributed by AV, 29-Jul-2023.)
⊢ (𝑥 = 𝑧 → (𝜑 ↔ 𝜓))    &   ⊢ (𝑦 = 𝑥 → (𝜓 ↔ 𝜒))    &   ⊢ (𝑧 = 𝑦 → (𝜒 ↔ 𝜑))    ⇒   ⊢ [𝑥⇄𝑦]𝜑
Theorem	ichan 47579	If two setvar variables are interchangeable in two wffs, then they are interchangeable in the conjunction of these two wffs. Notice that the reverse implication is not necessarily true. Corresponding theorems will hold for other commutative operations, too. (Contributed by AV, 31-Jul-2023.) Use df-ich 47570 instead of dfich2 47582 to reduce axioms. (Revised by SN, 4-May-2024.)
⊢ (([𝑎⇄𝑏]𝜑 ∧ [𝑎⇄𝑏]𝜓) → [𝑎⇄𝑏](𝜑 ∧ 𝜓))
Theorem	ichn 47580	Negation does not affect interchangeability. (Contributed by SN, 30-Aug-2023.)
⊢ ([𝑎⇄𝑏]𝜑 ↔ [𝑎⇄𝑏] ¬ 𝜑)
Theorem	ichim 47581	Formula building rule for implication in interchangeability. (Contributed by SN, 4-May-2024.)
⊢ (([𝑎⇄𝑏]𝜑 ∧ [𝑎⇄𝑏]𝜓) → [𝑎⇄𝑏](𝜑 → 𝜓))
Theorem	dfich2 47582*	Alternate definition of the property of a wff 𝜑 that the setvar variables 𝑥 and 𝑦 are interchangeable. (Contributed by AV and WL, 6-Aug-2023.)
⊢ ([𝑥⇄𝑦]𝜑 ↔ ∀𝑎∀𝑏([𝑎 / 𝑥][𝑏 / 𝑦]𝜑 ↔ [𝑏 / 𝑥][𝑎 / 𝑦]𝜑))
Theorem	ichcom 47583*	The interchangeability of setvar variables is commutative. (Contributed by AV, 20-Aug-2023.)
⊢ ([𝑥⇄𝑦]𝜓 ↔ [𝑦⇄𝑥]𝜓)
Theorem	ichbi12i 47584*	Equivalence for interchangeable setvar variables. (Contributed by AV, 29-Jul-2023.)
⊢ ((𝑥 = 𝑎 ∧ 𝑦 = 𝑏) → (𝜓 ↔ 𝜒))    ⇒   ⊢ ([𝑥⇄𝑦]𝜓 ↔ [𝑎⇄𝑏]𝜒)
Theorem	icheqid 47585	In an equality for the same setvar variable, the setvar variable is interchangeable by itself. Special case of ichid 47575 and icheq 47586 without distinct variables restriction. (Contributed by AV, 29-Jul-2023.)
⊢ [𝑥⇄𝑥]𝑥 = 𝑥
Theorem	icheq 47586*	In an equality of setvar variables, the setvar variables are interchangeable. (Contributed by AV, 29-Jul-2023.)
⊢ [𝑥⇄𝑦]𝑥 = 𝑦
Theorem	ichnfimlem 47587*	Lemma for ichnfim 47588: A substitution for a nonfree variable has no effect. (Contributed by Wolf Lammen, 6-Aug-2023.) Avoid ax-13 2374. (Revised by GG, 1-May-2024.)
⊢ (∀𝑦Ⅎ𝑥𝜑 → ([𝑎 / 𝑥][𝑏 / 𝑦]𝜑 ↔ [𝑏 / 𝑦]𝜑))
Theorem	ichnfim 47588*	If in an interchangeability context 𝑥 is not free in 𝜑, the same holds for 𝑦. (Contributed by Wolf Lammen, 6-Aug-2023.) (Revised by AV, 23-Sep-2023.)
⊢ ((∀𝑦Ⅎ𝑥𝜑 ∧ [𝑥⇄𝑦]𝜑) → ∀𝑥Ⅎ𝑦𝜑)
Theorem	ichnfb 47589*	If 𝑥 and 𝑦 are interchangeable in 𝜑, they are both free or both not free in 𝜑. (Contributed by Wolf Lammen, 6-Aug-2023.) (Revised by AV, 23-Sep-2023.)
⊢ ([𝑥⇄𝑦]𝜑 → (∀𝑥Ⅎ𝑦𝜑 ↔ ∀𝑦Ⅎ𝑥𝜑))
Theorem	ichal 47590*	Move a universal quantifier inside interchangeability. (Contributed by SN, 30-Aug-2023.)
⊢ (∀𝑥[𝑎⇄𝑏]𝜑 → [𝑎⇄𝑏]∀𝑥𝜑)
Theorem	ich2al 47591	Two setvar variables are always interchangeable when there are two universal quantifiers. (Contributed by SN, 23-Nov-2023.)
⊢ [𝑥⇄𝑦]∀𝑥∀𝑦𝜑
Theorem	ich2ex 47592	Two setvar variables are always interchangeable when there are two existential quantifiers. (Contributed by SN, 23-Nov-2023.)
⊢ [𝑥⇄𝑦]∃𝑥∃𝑦𝜑
Theorem	ichexmpl1 47593*	Example for interchangeable setvar variables in a statement of predicate calculus with equality. (Contributed by AV, 31-Jul-2023.)
⊢ [𝑎⇄𝑏]∃𝑎∃𝑏∃𝑐(𝑎 = 𝑏 ∧ 𝑎 ≠ 𝑐 ∧ 𝑏 ≠ 𝑐)
Theorem	ichexmpl2 47594*	Example for interchangeable setvar variables in an arithmetic expression. (Contributed by AV, 31-Jul-2023.)
⊢ [𝑎⇄𝑏]((𝑎 ∈ ℂ ∧ 𝑏 ∈ ℂ ∧ 𝑐 ∈ ℂ) → ((𝑎↑2) + (𝑏↑2)) = (𝑐↑2))
Theorem	ich2exprop 47595*	If the setvar variables are interchangeable in a wff, there is an ordered pair fulfilling the wff iff there is an unordered pair fulfilling the wff. (Contributed by AV, 16-Jul-2023.)
⊢ ((𝐴 ∈ 𝑋 ∧ 𝐵 ∈ 𝑋 ∧ [𝑎⇄𝑏]𝜑) → (∃𝑎∃𝑏({𝐴, 𝐵} = {𝑎, 𝑏} ∧ 𝜑) ↔ ∃𝑎∃𝑏(⟨𝐴, 𝐵⟩ = ⟨𝑎, 𝑏⟩ ∧ 𝜑)))
Theorem	ichnreuop 47596*	If the setvar variables are interchangeable in a wff, there is never a unique ordered pair with different components fulfilling the wff (because if ⟨𝑎, 𝑏⟩ fulfils the wff, then also ⟨𝑏, 𝑎⟩ fulfils the wff). (Contributed by AV, 27-Aug-2023.)
⊢ ([𝑎⇄𝑏]𝜑 → ¬ ∃!𝑝 ∈ (𝑋 × 𝑋)∃𝑎∃𝑏(𝑝 = ⟨𝑎, 𝑏⟩ ∧ 𝑎 ≠ 𝑏 ∧ 𝜑))
Theorem	ichreuopeq 47597*	If the setvar variables are interchangeable in a wff, and there is a unique ordered pair fulfilling the wff, then both setvar variables must be equal. (Contributed by AV, 28-Aug-2023.)
⊢ ([𝑎⇄𝑏]𝜑 → (∃!𝑝 ∈ (𝑋 × 𝑋)∃𝑎∃𝑏(𝑝 = ⟨𝑎, 𝑏⟩ ∧ 𝜑) → ∃𝑎∃𝑏(𝑎 = 𝑏 ∧ 𝜑)))
21.48.11.2  Set of unordered pairs
Theorem	sprid 47598	Two identical representations of the class of all unordered pairs. (Contributed by AV, 21-Nov-2021.)
⊢ {𝑝 ∣ ∃𝑎 ∈ V ∃𝑏 ∈ V 𝑝 = {𝑎, 𝑏}} = {𝑝 ∣ ∃𝑎∃𝑏 𝑝 = {𝑎, 𝑏}}
Theorem	elsprel 47599*	An unordered pair is an element of all unordered pairs. At least one of the two elements of the unordered pair must be a set. Otherwise, the unordered pair would be the empty set, see prprc 4719, which is not an element of all unordered pairs, see spr0nelg 47600. (Contributed by AV, 21-Nov-2021.)
⊢ ((𝐴 ∈ 𝑉 ∨ 𝐵 ∈ 𝑊) → {𝐴, 𝐵} ∈ {𝑝 ∣ ∃𝑎∃𝑏 𝑝 = {𝑎, 𝑏}})
Theorem	spr0nelg 47600*	The empty set is not an element of all unordered pairs. (Contributed by AV, 21-Nov-2021.)
⊢ ∅ ∉ {𝑝 ∣ ∃𝑎∃𝑏 𝑝 = {𝑎, 𝑏}}