Home | Metamath
Proof Explorer Theorem List (p. 447 of 464) | < Previous Next > |
Bad symbols? Try the
GIF version. |
||
Mirrors > Metamath Home Page > MPE Home Page > Theorem List Contents > Recent Proofs This page: Page List |
Color key: | Metamath Proof Explorer
(1-29181) |
Hilbert Space Explorer
(29182-30704) |
Users' Mathboxes
(30705-46395) |
Type | Label | Description |
---|---|---|
Statement | ||
Theorem | ndmafv2nrn 44601 | The value of a class outside its domain is not in the range, compare with ndmfv 6786. (Contributed by AV, 2-Sep-2022.) |
⊢ (¬ 𝐴 ∈ dom 𝐹 → (𝐹''''𝐴) ∉ ran 𝐹) | ||
Theorem | funressndmafv2rn 44602 | The alternate function value at a class 𝐴 is defined, i.e., in the range of the function if the function is defined at 𝐴. (Contributed by AV, 2-Sep-2022.) |
⊢ (𝐹 defAt 𝐴 → (𝐹''''𝐴) ∈ ran 𝐹) | ||
Theorem | afv2ndefb 44603 | Two ways to say that an alternate function value is not defined. (Contributed by AV, 5-Sep-2022.) |
⊢ ((𝐹''''𝐴) = 𝒫 ∪ ran 𝐹 ↔ (𝐹''''𝐴) ∉ ran 𝐹) | ||
Theorem | nfunsnafv2 44604 | If the restriction of a class to a singleton is not a function, its value at the singleton element is undefined, compare with nfunsn 6793. (Contributed by AV, 2-Sep-2022.) |
⊢ (¬ Fun (𝐹 ↾ {𝐴}) → (𝐹''''𝐴) ∉ ran 𝐹) | ||
Theorem | afv2prc 44605 | A function's value at a proper class is not defined, compare with fvprc 6748. (Contributed by AV, 5-Sep-2022.) |
⊢ (¬ 𝐴 ∈ V → (𝐹''''𝐴) ∉ ran 𝐹) | ||
Theorem | dfatafv2rnb 44606 | The alternate function value at a class 𝐴 is defined, i.e. in the range of the function, iff the function is defined at 𝐴. (Contributed by AV, 2-Sep-2022.) |
⊢ (𝐹 defAt 𝐴 ↔ (𝐹''''𝐴) ∈ ran 𝐹) | ||
Theorem | afv2orxorb 44607 | If a set is in the range of a function, the alternate function value at a class 𝐴 equals this set or is not in the range of the function iff the alternate function value at the class 𝐴 either equals this set or is not in the range of the function. If 𝐵 ∉ ran 𝐹, both disjuncts of the exclusive or can be true: (𝐹''''𝐴) = 𝐵 → (𝐹''''𝐴) ∉ ran 𝐹. (Contributed by AV, 11-Sep-2022.) |
⊢ (𝐵 ∈ ran 𝐹 → (((𝐹''''𝐴) = 𝐵 ∨ (𝐹''''𝐴) ∉ ran 𝐹) ↔ ((𝐹''''𝐴) = 𝐵 ⊻ (𝐹''''𝐴) ∉ ran 𝐹))) | ||
Theorem | dmafv2rnb 44608 | The alternate function value at a class 𝐴 is defined, i.e., in the range of the function, iff 𝐴 is in the domain of the function. (Contributed by AV, 3-Sep-2022.) |
⊢ (Fun (𝐹 ↾ {𝐴}) → (𝐴 ∈ dom 𝐹 ↔ (𝐹''''𝐴) ∈ ran 𝐹)) | ||
Theorem | fundmafv2rnb 44609 | The alternate function value at a class 𝐴 is defined, i.e., in the range of the function iff 𝐴 is in the domain of the function. (Contributed by AV, 3-Sep-2022.) |
⊢ (Fun 𝐹 → (𝐴 ∈ dom 𝐹 ↔ (𝐹''''𝐴) ∈ ran 𝐹)) | ||
Theorem | afv2elrn 44610 | An alternate function value belongs to the range of the function, analogous to fvelrn 6936. (Contributed by AV, 3-Sep-2022.) |
⊢ ((Fun 𝐹 ∧ 𝐴 ∈ dom 𝐹) → (𝐹''''𝐴) ∈ ran 𝐹) | ||
Theorem | afv20defat 44611 | If the alternate function value at an argument is the empty set, the function is defined at this argument. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹''''𝐴) = ∅ → 𝐹 defAt 𝐴) | ||
Theorem | fnafv2elrn 44612 | An alternate function value belongs to the range of the function, analogous to fnfvelrn 6940. (Contributed by AV, 2-Sep-2022.) |
⊢ ((𝐹 Fn 𝐴 ∧ 𝐵 ∈ 𝐴) → (𝐹''''𝐵) ∈ ran 𝐹) | ||
Theorem | fafv2elrn 44613 | An alternate function value belongs to the codomain of the function, analogous to ffvelrn 6941. (Contributed by AV, 2-Sep-2022.) |
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐶 ∈ 𝐴) → (𝐹''''𝐶) ∈ 𝐵) | ||
Theorem | fafv2elrnb 44614 | An alternate function value is defined, i.e., belongs to the range of the function, iff its argument is in the domain of the function. (Contributed by AV, 3-Sep-2022.) |
⊢ (𝐹:𝐴⟶𝐵 → (𝐶 ∈ 𝐴 ↔ (𝐹''''𝐶) ∈ ran 𝐹)) | ||
Theorem | frnvafv2v 44615 | If the codomain of a function is a set, the alternate function value is always also a set. (Contributed by AV, 4-Sep-2022.) |
⊢ ((𝐹:𝐴⟶𝐵 ∧ 𝐵 ∈ 𝑉) → (𝐹''''𝐶) ∈ V) | ||
Theorem | tz6.12-2-afv2 44616* | Function value when 𝐹 is (locally) not a function. Theorem 6.12(2) of [TakeutiZaring] p. 27, analogous to tz6.12-2 6745. (Contributed by AV, 5-Sep-2022.) |
⊢ (¬ ∃!𝑥 𝐴𝐹𝑥 → (𝐹''''𝐴) ∉ ran 𝐹) | ||
Theorem | afv2eu 44617* | The value of a function at a unique point, analogous to fveu 6746. (Contributed by AV, 5-Sep-2022.) |
⊢ (∃!𝑥 𝐴𝐹𝑥 → (𝐹''''𝐴) = ∪ {𝑥 ∣ 𝐴𝐹𝑥}) | ||
Theorem | afv2res 44618 | The value of a restricted function for an argument at which the function is defined. Analog to fvres 6775. (Contributed by AV, 5-Sep-2022.) |
⊢ ((𝐹 defAt 𝐴 ∧ 𝐴 ∈ 𝐵) → ((𝐹 ↾ 𝐵)''''𝐴) = (𝐹''''𝐴)) | ||
Theorem | tz6.12-afv2 44619* | Function value (Theorem 6.12(1) of [TakeutiZaring] p. 27), analogous to tz6.12 6779. (Contributed by AV, 5-Sep-2022.) |
⊢ ((〈𝐴, 𝑦〉 ∈ 𝐹 ∧ ∃!𝑦〈𝐴, 𝑦〉 ∈ 𝐹) → (𝐹''''𝐴) = 𝑦) | ||
Theorem | tz6.12-1-afv2 44620* | Function value (Theorem 6.12(1) of [TakeutiZaring] p. 27), analogous to tz6.12-1 6778. (Contributed by AV, 5-Sep-2022.) |
⊢ ((𝐴𝐹𝑦 ∧ ∃!𝑦 𝐴𝐹𝑦) → (𝐹''''𝐴) = 𝑦) | ||
Theorem | tz6.12c-afv2 44621* | Corollary of Theorem 6.12(1) of [TakeutiZaring] p. 27, analogous to tz6.12c 6781. (Contributed by AV, 5-Sep-2022.) |
⊢ (∃!𝑦 𝐴𝐹𝑦 → ((𝐹''''𝐴) = 𝑦 ↔ 𝐴𝐹𝑦)) | ||
Theorem | tz6.12i-afv2 44622 | Corollary of Theorem 6.12(2) of [TakeutiZaring] p. 27. analogous to tz6.12i 6782. (Contributed by AV, 5-Sep-2022.) |
⊢ (𝐵 ∈ ran 𝐹 → ((𝐹''''𝐴) = 𝐵 → 𝐴𝐹𝐵)) | ||
Theorem | funressnbrafv2 44623 | The second argument of a binary relation on a function is the function's value, analogous to funbrfv 6802. (Contributed by AV, 7-Sep-2022.) |
⊢ (((𝐴 ∈ 𝑉 ∧ 𝐵 ∈ 𝑊) ∧ Fun (𝐹 ↾ {𝐴})) → (𝐴𝐹𝐵 → (𝐹''''𝐴) = 𝐵)) | ||
Theorem | dfatbrafv2b 44624 | Equivalence of function value and binary relation, analogous to fnbrfvb 6804 or funbrfvb 6806. 𝐵 ∈ V is required, because otherwise 𝐴𝐹𝐵 ↔ ∅ ∈ 𝐹 can be true, but (𝐹''''𝐴) = 𝐵 is always false (because of dfatafv2ex 44592). (Contributed by AV, 6-Sep-2022.) |
⊢ ((𝐹 defAt 𝐴 ∧ 𝐵 ∈ 𝑊) → ((𝐹''''𝐴) = 𝐵 ↔ 𝐴𝐹𝐵)) | ||
Theorem | dfatopafv2b 44625 | Equivalence of function value and ordered pair membership, analogous to fnopfvb 6805 or funopfvb 6807. (Contributed by AV, 6-Sep-2022.) |
⊢ ((𝐹 defAt 𝐴 ∧ 𝐵 ∈ 𝑊) → ((𝐹''''𝐴) = 𝐵 ↔ 〈𝐴, 𝐵〉 ∈ 𝐹)) | ||
Theorem | funbrafv2 44626 | The second argument of a binary relation on a function is the function's value, analogous to funbrfv 6802. (Contributed by AV, 6-Sep-2022.) |
⊢ (Fun 𝐹 → (𝐴𝐹𝐵 → (𝐹''''𝐴) = 𝐵)) | ||
Theorem | fnbrafv2b 44627 | Equivalence of function value and binary relation, analogous to fnbrfvb 6804. (Contributed by AV, 6-Sep-2022.) |
⊢ ((𝐹 Fn 𝐴 ∧ 𝐵 ∈ 𝐴) → ((𝐹''''𝐵) = 𝐶 ↔ 𝐵𝐹𝐶)) | ||
Theorem | fnopafv2b 44628 | Equivalence of function value and ordered pair membership, analogous to fnopfvb 6805. (Contributed by AV, 6-Sep-2022.) |
⊢ ((𝐹 Fn 𝐴 ∧ 𝐵 ∈ 𝐴) → ((𝐹''''𝐵) = 𝐶 ↔ 〈𝐵, 𝐶〉 ∈ 𝐹)) | ||
Theorem | funbrafv22b 44629 | Equivalence of function value and binary relation, analogous to funbrfvb 6806. (Contributed by AV, 6-Sep-2022.) |
⊢ ((Fun 𝐹 ∧ 𝐴 ∈ dom 𝐹) → ((𝐹''''𝐴) = 𝐵 ↔ 𝐴𝐹𝐵)) | ||
Theorem | funopafv2b 44630 | Equivalence of function value and ordered pair membership, analogous to funopfvb 6807. (Contributed by AV, 6-Sep-2022.) |
⊢ ((Fun 𝐹 ∧ 𝐴 ∈ dom 𝐹) → ((𝐹''''𝐴) = 𝐵 ↔ 〈𝐴, 𝐵〉 ∈ 𝐹)) | ||
Theorem | dfatsnafv2 44631 | Singleton of function value, analogous to fnsnfv 6829. (Contributed by AV, 7-Sep-2022.) |
⊢ (𝐹 defAt 𝐴 → {(𝐹''''𝐴)} = (𝐹 “ {𝐴})) | ||
Theorem | dfafv23 44632* | A definition of function value in terms of iota, analogous to dffv3 6752. (Contributed by AV, 6-Sep-2022.) |
⊢ (𝐹 defAt 𝐴 → (𝐹''''𝐴) = (℩𝑥𝑥 ∈ (𝐹 “ {𝐴}))) | ||
Theorem | dfatdmfcoafv2 44633 | Domain of a function composition, analogous to dmfco 6846. (Contributed by AV, 7-Sep-2022.) |
⊢ (𝐺 defAt 𝐴 → (𝐴 ∈ dom (𝐹 ∘ 𝐺) ↔ (𝐺''''𝐴) ∈ dom 𝐹)) | ||
Theorem | dfatcolem 44634* | Lemma for dfatco 44635. (Contributed by AV, 8-Sep-2022.) |
⊢ ((𝐺 defAt 𝑋 ∧ 𝐹 defAt (𝐺''''𝑋)) → ∃!𝑦 𝑋(𝐹 ∘ 𝐺)𝑦) | ||
Theorem | dfatco 44635 | The predicate "defined at" for a function composition. (Contributed by AV, 8-Sep-2022.) |
⊢ ((𝐺 defAt 𝑋 ∧ 𝐹 defAt (𝐺''''𝑋)) → (𝐹 ∘ 𝐺) defAt 𝑋) | ||
Theorem | afv2co2 44636 | Value of a function composition, analogous to fvco2 6847. (Contributed by AV, 8-Sep-2022.) |
⊢ ((𝐺 defAt 𝑋 ∧ 𝐹 defAt (𝐺''''𝑋)) → ((𝐹 ∘ 𝐺)''''𝑋) = (𝐹''''(𝐺''''𝑋))) | ||
Theorem | rlimdmafv2 44637 | Two ways to express that a function has a limit, analogous to rlimdm 15188. (Contributed by AV, 5-Sep-2022.) |
⊢ (𝜑 → 𝐹:𝐴⟶ℂ) & ⊢ (𝜑 → sup(𝐴, ℝ*, < ) = +∞) ⇒ ⊢ (𝜑 → (𝐹 ∈ dom ⇝𝑟 ↔ 𝐹 ⇝𝑟 ( ⇝𝑟 ''''𝐹))) | ||
Theorem | dfafv22 44638 | Alternate definition of (𝐹''''𝐴) using (𝐹‘𝐴) directly. (Contributed by AV, 3-Sep-2022.) |
⊢ (𝐹''''𝐴) = if(𝐹 defAt 𝐴, (𝐹‘𝐴), 𝒫 ∪ ran 𝐹) | ||
Theorem | afv2ndeffv0 44639 | If the alternate function value at an argument is undefined, i.e., not in the range of the function, the function's value at this argument is the empty set. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹''''𝐴) ∉ ran 𝐹 → (𝐹‘𝐴) = ∅) | ||
Theorem | dfatafv2eqfv 44640 | If a function is defined at a class 𝐴, the alternate function value equals the function's value at 𝐴. (Contributed by AV, 3-Sep-2022.) |
⊢ (𝐹 defAt 𝐴 → (𝐹''''𝐴) = (𝐹‘𝐴)) | ||
Theorem | afv2rnfveq 44641 | If the alternate function value is defined, i.e., in the range of the function, the alternate function value equals the function's value. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹''''𝐴) ∈ ran 𝐹 → (𝐹''''𝐴) = (𝐹‘𝐴)) | ||
Theorem | afv20fv0 44642 | If the alternate function value at an argument is the empty set, the function's value at this argument is the empty set. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹''''𝐴) = ∅ → (𝐹‘𝐴) = ∅) | ||
Theorem | afv2fvn0fveq 44643 | If the function's value at an argument is not the empty set, it equals the alternate function value at this argument. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹‘𝐴) ≠ ∅ → (𝐹''''𝐴) = (𝐹‘𝐴)) | ||
Theorem | afv2fv0 44644 | If the function's value at an argument is the empty set, then the alternate function value at this argument is the empty set or undefined. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹‘𝐴) = ∅ → ((𝐹''''𝐴) = ∅ ∨ (𝐹''''𝐴) ∉ ran 𝐹)) | ||
Theorem | afv2fv0b 44645 | The function's value at an argument is the empty set if and only if the alternate function value at this argument is the empty set or undefined. (Contributed by AV, 3-Sep-2022.) |
⊢ ((𝐹‘𝐴) = ∅ ↔ ((𝐹''''𝐴) = ∅ ∨ (𝐹''''𝐴) ∉ ran 𝐹)) | ||
Theorem | afv2fv0xorb 44646 | If a set is in the range of a function, the function's value at an argument is the empty set if and only if the alternate function value at this argument is either the empty set or undefined. (Contributed by AV, 11-Sep-2022.) |
⊢ (∅ ∈ ran 𝐹 → ((𝐹‘𝐴) = ∅ ↔ ((𝐹''''𝐴) = ∅ ⊻ (𝐹''''𝐴) ∉ ran 𝐹))) | ||
Theorem | an4com24 44647 | Rearrangement of 4 conjuncts: second and forth positions interchanged. (Contributed by AV, 18-Feb-2022.) |
⊢ (((𝜑 ∧ 𝜓) ∧ (𝜒 ∧ 𝜃)) ↔ ((𝜑 ∧ 𝜃) ∧ (𝜒 ∧ 𝜓))) | ||
Theorem | 3an4ancom24 44648 | Commutative law for a conjunction with a triple conjunction: second and forth positions interchanged. (Contributed by AV, 18-Feb-2022.) |
⊢ (((𝜑 ∧ 𝜓 ∧ 𝜒) ∧ 𝜃) ↔ ((𝜑 ∧ 𝜃 ∧ 𝜒) ∧ 𝜓)) | ||
Theorem | 4an21 44649 | Rearrangement of 4 conjuncts with a triple conjunction. (Contributed by AV, 4-Mar-2022.) |
⊢ (((𝜑 ∧ 𝜓) ∧ 𝜒 ∧ 𝜃) ↔ (𝜓 ∧ (𝜑 ∧ 𝜒 ∧ 𝜃))) | ||
Syntax | cnelbr 44650 | Extend wff notation to include the 'not elemet of' relation. |
class _∉ | ||
Definition | df-nelbr 44651* | Define negated membership as binary relation. Analogous to df-eprel 5486 (the membership relation). (Contributed by AV, 26-Dec-2021.) |
⊢ _∉ = {〈𝑥, 𝑦〉 ∣ ¬ 𝑥 ∈ 𝑦} | ||
Theorem | dfnelbr2 44652 | Alternate definition of the negated membership as binary relation. (Proposed by BJ, 27-Dec-2021.) (Contributed by AV, 27-Dec-2021.) |
⊢ _∉ = ((V × V) ∖ E ) | ||
Theorem | nelbr 44653 | The binary relation of a set not being a member of another set. (Contributed by AV, 26-Dec-2021.) |
⊢ ((𝐴 ∈ 𝑉 ∧ 𝐵 ∈ 𝑊) → (𝐴 _∉ 𝐵 ↔ ¬ 𝐴 ∈ 𝐵)) | ||
Theorem | nelbrim 44654 | If a set is related to another set by the negated membership relation, then it is not a member of the other set. The other direction of the implication is not generally true, because if 𝐴 is a proper class, then ¬ 𝐴 ∈ 𝐵 would be true, but not 𝐴 _∉ 𝐵. (Contributed by AV, 26-Dec-2021.) |
⊢ (𝐴 _∉ 𝐵 → ¬ 𝐴 ∈ 𝐵) | ||
Theorem | nelbrnel 44655 | A set is related to another set by the negated membership relation iff it is not a member of the other set. (Contributed by AV, 26-Dec-2021.) |
⊢ ((𝐴 ∈ 𝑉 ∧ 𝐵 ∈ 𝑊) → (𝐴 _∉ 𝐵 ↔ 𝐴 ∉ 𝐵)) | ||
Theorem | nelbrnelim 44656 | If a set is related to another set by the negated membership relation, then it is not a member of the other set. (Contributed by AV, 26-Dec-2021.) |
⊢ (𝐴 _∉ 𝐵 → 𝐴 ∉ 𝐵) | ||
Theorem | ralralimp 44657* | Selecting one of two alternatives within a restricted generalization if one of the alternatives is false. (Contributed by AV, 6-Sep-2018.) (Proof shortened by AV, 13-Oct-2018.) |
⊢ ((𝜑 ∧ 𝐴 ≠ ∅) → (∀𝑥 ∈ 𝐴 ((𝜑 → (𝜃 ∨ 𝜏)) ∧ ¬ 𝜃) → 𝜏)) | ||
Theorem | otiunsndisjX 44658* | The union of singletons consisting of ordered triples which have distinct first and third components are disjunct. (Contributed by Alexander van der Vekens, 10-Mar-2018.) |
⊢ (𝐵 ∈ 𝑋 → Disj 𝑎 ∈ 𝑉 ∪ 𝑐 ∈ 𝑊 {〈𝑎, 𝐵, 𝑐〉}) | ||
Theorem | fvifeq 44659 | Equality of function values with conditional arguments, see also fvif 6772. (Contributed by Alexander van der Vekens, 21-May-2018.) |
⊢ (𝐴 = if(𝜑, 𝐵, 𝐶) → (𝐹‘𝐴) = if(𝜑, (𝐹‘𝐵), (𝐹‘𝐶))) | ||
Theorem | rnfdmpr 44660 | The range of a one-to-one function 𝐹 of an unordered pair into a set is the unordered pair of the function values. (Contributed by Alexander van der Vekens, 2-Feb-2018.) |
⊢ ((𝑋 ∈ 𝑉 ∧ 𝑌 ∈ 𝑊) → (𝐹 Fn {𝑋, 𝑌} → ran 𝐹 = {(𝐹‘𝑋), (𝐹‘𝑌)})) | ||
Theorem | imarnf1pr 44661 | The image of the range of a function 𝐹 under a function 𝐸 if 𝐹 is a function from a pair into the domain of 𝐸. (Contributed by Alexander van der Vekens, 2-Feb-2018.) |
⊢ ((𝑋 ∈ 𝑉 ∧ 𝑌 ∈ 𝑊) → (((𝐹:{𝑋, 𝑌}⟶dom 𝐸 ∧ 𝐸:dom 𝐸⟶𝑅) ∧ ((𝐸‘(𝐹‘𝑋)) = 𝐴 ∧ (𝐸‘(𝐹‘𝑌)) = 𝐵)) → (𝐸 “ ran 𝐹) = {𝐴, 𝐵})) | ||
Theorem | funop1 44662* | A function is an ordered pair iff it is a singleton of an ordered pair. (Contributed by AV, 20-Sep-2020.) (Avoid depending on this detail.) |
⊢ (∃𝑥∃𝑦 𝐹 = 〈𝑥, 𝑦〉 → (Fun 𝐹 ↔ ∃𝑥∃𝑦 𝐹 = {〈𝑥, 𝑦〉})) | ||
Theorem | fun2dmnopgexmpl 44663 | A function with a domain containing (at least) two different elements is not an ordered pair. (Contributed by AV, 21-Sep-2020.) (Avoid depending on this detail.) |
⊢ (𝐺 = {〈0, 1〉, 〈1, 1〉} → ¬ 𝐺 ∈ (V × V)) | ||
Theorem | opabresex0d 44664* | A collection of ordered pairs, the class of all possible second components being a set, with a restriction of a binary relation is a set. (Contributed by Alexander van der Vekens, 1-Nov-2017.) (Revised by AV, 1-Jan-2021.) |
⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑥 ∈ 𝐶) & ⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝜃) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → {𝑦 ∣ 𝜃} ∈ 𝑉) & ⊢ (𝜑 → 𝐶 ∈ 𝑊) ⇒ ⊢ (𝜑 → {〈𝑥, 𝑦〉 ∣ (𝑥𝑅𝑦 ∧ 𝜓)} ∈ V) | ||
Theorem | opabbrfex0d 44665* | A collection of ordered pairs, the class of all possible second components being a set, is a set. (Contributed by AV, 15-Jan-2021.) |
⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑥 ∈ 𝐶) & ⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝜃) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → {𝑦 ∣ 𝜃} ∈ 𝑉) & ⊢ (𝜑 → 𝐶 ∈ 𝑊) ⇒ ⊢ (𝜑 → {〈𝑥, 𝑦〉 ∣ 𝑥𝑅𝑦} ∈ V) | ||
Theorem | opabresexd 44666* | A collection of ordered pairs, the second component being a function, with a restriction of a binary relation is a set. (Contributed by Alexander van der Vekens, 1-Nov-2017.) (Revised by AV, 15-Jan-2021.) |
⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑥 ∈ 𝐶) & ⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑦:𝐴⟶𝐵) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → 𝐴 ∈ 𝑈) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → 𝐵 ∈ 𝑉) & ⊢ (𝜑 → 𝐶 ∈ 𝑊) ⇒ ⊢ (𝜑 → {〈𝑥, 𝑦〉 ∣ (𝑥𝑅𝑦 ∧ 𝜓)} ∈ V) | ||
Theorem | opabbrfexd 44667* | A collection of ordered pairs, the second component being a function, is a set. (Contributed by AV, 15-Jan-2021.) |
⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑥 ∈ 𝐶) & ⊢ ((𝜑 ∧ 𝑥𝑅𝑦) → 𝑦:𝐴⟶𝐵) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → 𝐴 ∈ 𝑈) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐶) → 𝐵 ∈ 𝑉) & ⊢ (𝜑 → 𝐶 ∈ 𝑊) ⇒ ⊢ (𝜑 → {〈𝑥, 𝑦〉 ∣ 𝑥𝑅𝑦} ∈ V) | ||
Theorem | f1oresf1orab 44668* | Build a bijection by restricting the domain of a bijection. (Contributed by AV, 1-Aug-2022.) |
⊢ 𝐹 = (𝑥 ∈ 𝐴 ↦ 𝐶) & ⊢ (𝜑 → 𝐹:𝐴–1-1-onto→𝐵) & ⊢ (𝜑 → 𝐷 ⊆ 𝐴) & ⊢ ((𝜑 ∧ 𝑥 ∈ 𝐴 ∧ 𝑦 = 𝐶) → (𝜒 ↔ 𝑥 ∈ 𝐷)) ⇒ ⊢ (𝜑 → (𝐹 ↾ 𝐷):𝐷–1-1-onto→{𝑦 ∈ 𝐵 ∣ 𝜒}) | ||
Theorem | f1oresf1o 44669* | Build a bijection by restricting the domain of a bijection. (Contributed by AV, 31-Jul-2022.) |
⊢ (𝜑 → 𝐹:𝐴–1-1-onto→𝐵) & ⊢ (𝜑 → 𝐷 ⊆ 𝐴) & ⊢ (𝜑 → (∃𝑥 ∈ 𝐷 (𝐹‘𝑥) = 𝑦 ↔ (𝑦 ∈ 𝐵 ∧ 𝜒))) ⇒ ⊢ (𝜑 → (𝐹 ↾ 𝐷):𝐷–1-1-onto→{𝑦 ∈ 𝐵 ∣ 𝜒}) | ||
Theorem | f1oresf1o2 44670* | Build a bijection by restricting the domain of a bijection. (Contributed by AV, 31-Jul-2022.) |
⊢ (𝜑 → 𝐹:𝐴–1-1-onto→𝐵) & ⊢ (𝜑 → 𝐷 ⊆ 𝐴) & ⊢ ((𝜑 ∧ 𝑦 = (𝐹‘𝑥)) → (𝑥 ∈ 𝐷 ↔ 𝜒)) ⇒ ⊢ (𝜑 → (𝐹 ↾ 𝐷):𝐷–1-1-onto→{𝑦 ∈ 𝐵 ∣ 𝜒}) | ||
Theorem | fvmptrab 44671* | Value of a function mapping a set to a class abstraction restricting a class depending on the argument of the function. More general version of fvmptrabfv 6888, but relying on the fact that out-of-domain arguments evaluate to the empty set, which relies on set.mm's particular encoding. (Contributed by AV, 14-Feb-2022.) |
⊢ 𝐹 = (𝑥 ∈ 𝑉 ↦ {𝑦 ∈ 𝑀 ∣ 𝜑}) & ⊢ (𝑥 = 𝑋 → (𝜑 ↔ 𝜓)) & ⊢ (𝑥 = 𝑋 → 𝑀 = 𝑁) & ⊢ (𝑋 ∈ 𝑉 → 𝑁 ∈ V) & ⊢ (𝑋 ∉ 𝑉 → 𝑁 = ∅) ⇒ ⊢ (𝐹‘𝑋) = {𝑦 ∈ 𝑁 ∣ 𝜓} | ||
Theorem | fvmptrabdm 44672* | Value of a function mapping a set to a class abstraction restricting the value of another function. See also fvmptrabfv 6888. (Suggested by BJ, 18-Feb-2022.) (Contributed by AV, 18-Feb-2022.) |
⊢ 𝐹 = (𝑥 ∈ 𝑉 ↦ {𝑦 ∈ (𝐺‘𝑌) ∣ 𝜑}) & ⊢ (𝑥 = 𝑋 → (𝜑 ↔ 𝜓)) & ⊢ (𝑌 ∈ dom 𝐺 → 𝑋 ∈ dom 𝐹) ⇒ ⊢ (𝐹‘𝑋) = {𝑦 ∈ (𝐺‘𝑌) ∣ 𝜓} | ||
Theorem | leltletr 44673 | Transitive law, weaker form of lelttr 10996. (Contributed by AV, 14-Oct-2018.) |
⊢ ((𝐴 ∈ ℝ ∧ 𝐵 ∈ ℝ ∧ 𝐶 ∈ ℝ) → ((𝐴 ≤ 𝐵 ∧ 𝐵 < 𝐶) → 𝐴 ≤ 𝐶)) | ||
Theorem | cnambpcma 44674 | ((a-b)+c)-a = c-a holds for complex numbers a,b,c. (Contributed by Alexander van der Vekens, 23-Mar-2018.) |
⊢ ((𝐴 ∈ ℂ ∧ 𝐵 ∈ ℂ ∧ 𝐶 ∈ ℂ) → (((𝐴 − 𝐵) + 𝐶) − 𝐴) = (𝐶 − 𝐵)) | ||
Theorem | cnapbmcpd 44675 | ((a+b)-c)+d = ((a+d)+b)-c holds for complex numbers a,b,c,d. (Contributed by Alexander van der Vekens, 23-Mar-2018.) |
⊢ (((𝐴 ∈ ℂ ∧ 𝐵 ∈ ℂ) ∧ (𝐶 ∈ ℂ ∧ 𝐷 ∈ ℂ)) → (((𝐴 + 𝐵) − 𝐶) + 𝐷) = (((𝐴 + 𝐷) + 𝐵) − 𝐶)) | ||
Theorem | addsubeq0 44676 | The sum of two complex numbers is equal to the difference of these two complex numbers iff the subtrahend is 0. (Contributed by AV, 8-May-2023.) |
⊢ ((𝐴 ∈ ℂ ∧ 𝐵 ∈ ℂ) → ((𝐴 + 𝐵) = (𝐴 − 𝐵) ↔ 𝐵 = 0)) | ||
Theorem | leaddsuble 44677 | Addition and subtraction on one side of "less than or equal to". (Contributed by Alexander van der Vekens, 18-Mar-2018.) |
⊢ ((𝐴 ∈ ℝ ∧ 𝐵 ∈ ℝ ∧ 𝐶 ∈ ℝ) → (𝐵 ≤ 𝐶 ↔ ((𝐴 + 𝐵) − 𝐶) ≤ 𝐴)) | ||
Theorem | 2leaddle2 44678 | If two real numbers are less than a third real number, the sum of the real numbers is less than twice the third real number. (Contributed by Alexander van der Vekens, 21-May-2018.) |
⊢ ((𝐴 ∈ ℝ ∧ 𝐵 ∈ ℝ ∧ 𝐶 ∈ ℝ) → ((𝐴 < 𝐶 ∧ 𝐵 < 𝐶) → (𝐴 + 𝐵) < (2 · 𝐶))) | ||
Theorem | ltnltne 44679 | Variant of trichotomy law for 'less than'. (Contributed by Alexander van der Vekens, 8-Jun-2018.) |
⊢ ((𝐴 ∈ ℝ ∧ 𝐵 ∈ ℝ) → (𝐴 < 𝐵 ↔ (¬ 𝐵 < 𝐴 ∧ ¬ 𝐵 = 𝐴))) | ||
Theorem | p1lep2 44680 | A real number increasd by 1 is less than or equal to the number increased by 2. (Contributed by Alexander van der Vekens, 17-Sep-2018.) |
⊢ (𝑁 ∈ ℝ → (𝑁 + 1) ≤ (𝑁 + 2)) | ||
Theorem | ltsubsubaddltsub 44681 | If the result of subtracting two numbers is greater than a number, the result of adding one of these subtracted numbers to the number is less than the result of subtracting the other subtracted number only. (Contributed by Alexander van der Vekens, 9-Jun-2018.) |
⊢ ((𝐽 ∈ ℝ ∧ (𝐿 ∈ ℝ ∧ 𝑀 ∈ ℝ ∧ 𝑁 ∈ ℝ)) → (𝐽 < ((𝐿 − 𝑀) − 𝑁) ↔ (𝐽 + 𝑀) < (𝐿 − 𝑁))) | ||
Theorem | zm1nn 44682 | An integer minus 1 is positive under certain circumstances. (Contributed by Alexander van der Vekens, 9-Jun-2018.) |
⊢ ((𝑁 ∈ ℕ0 ∧ 𝐿 ∈ ℤ) → ((𝐽 ∈ ℝ ∧ 0 ≤ 𝐽 ∧ 𝐽 < ((𝐿 − 𝑁) − 1)) → (𝐿 − 1) ∈ ℕ)) | ||
Theorem | readdcnnred 44683 | The sum of a real number and an imaginary number is not a real number. (Contributed by AV, 23-Jan-2023.) |
⊢ (𝜑 → 𝐴 ∈ ℝ) & ⊢ (𝜑 → 𝐵 ∈ (ℂ ∖ ℝ)) ⇒ ⊢ (𝜑 → (𝐴 + 𝐵) ∉ ℝ) | ||
Theorem | resubcnnred 44684 | The difference of a real number and an imaginary number is not a real number. (Contributed by AV, 23-Jan-2023.) |
⊢ (𝜑 → 𝐴 ∈ ℝ) & ⊢ (𝜑 → 𝐵 ∈ (ℂ ∖ ℝ)) ⇒ ⊢ (𝜑 → (𝐴 − 𝐵) ∉ ℝ) | ||
Theorem | recnmulnred 44685 | The product of a real number and an imaginary number is not a real number. (Contributed by AV, 23-Jan-2023.) |
⊢ (𝜑 → 𝐴 ∈ ℝ) & ⊢ (𝜑 → 𝐵 ∈ (ℂ ∖ ℝ)) & ⊢ (𝜑 → 𝐴 ≠ 0) ⇒ ⊢ (𝜑 → (𝐴 · 𝐵) ∉ ℝ) | ||
Theorem | cndivrenred 44686 | The quotient of an imaginary number and a real number is not a real number. (Contributed by AV, 23-Jan-2023.) |
⊢ (𝜑 → 𝐴 ∈ ℝ) & ⊢ (𝜑 → 𝐵 ∈ (ℂ ∖ ℝ)) & ⊢ (𝜑 → 𝐴 ≠ 0) ⇒ ⊢ (𝜑 → (𝐵 / 𝐴) ∉ ℝ) | ||
Theorem | sqrtnegnre 44687 | The square root of a negative number is not a real number. (Contributed by AV, 28-Feb-2023.) |
⊢ ((𝑋 ∈ ℝ ∧ 𝑋 < 0) → (√‘𝑋) ∉ ℝ) | ||
Theorem | nn0resubcl 44688 | Closure law for subtraction of reals, restricted to nonnegative integers. (Contributed by Alexander van der Vekens, 6-Apr-2018.) |
⊢ ((𝐴 ∈ ℕ0 ∧ 𝐵 ∈ ℕ0) → (𝐴 − 𝐵) ∈ ℝ) | ||
Theorem | zgeltp1eq 44689 | If an integer is between another integer and its successor, the integer is equal to the other integer. (Contributed by AV, 30-May-2020.) |
⊢ ((𝐼 ∈ ℤ ∧ 𝐴 ∈ ℤ) → ((𝐴 ≤ 𝐼 ∧ 𝐼 < (𝐴 + 1)) → 𝐼 = 𝐴)) | ||
Theorem | 1t10e1p1e11 44690 | 11 is 1 times 10 to the power of 1, plus 1. (Contributed by AV, 4-Aug-2020.) (Revised by AV, 9-Sep-2021.) |
⊢ ;11 = ((1 · (;10↑1)) + 1) | ||
Theorem | deccarry 44691 | Add 1 to a 2 digit number with carry. This is a special case of decsucc 12407, but in closed form. As observed by ML, this theorem allows for carrying the 1 down multiple decimal constructors, so we can carry the 1 multiple times down a multi-digit number, e.g., by applying this theorem three times we get (;;999 + 1) = ;;;1000. (Contributed by AV, 4-Aug-2020.) (Revised by ML, 8-Aug-2020.) (Proof shortened by AV, 10-Sep-2021.) |
⊢ (𝐴 ∈ ℕ → (;𝐴9 + 1) = ;(𝐴 + 1)0) | ||
Theorem | eluzge0nn0 44692 | If an integer is greater than or equal to a nonnegative integer, then it is a nonnegative integer. (Contributed by Alexander van der Vekens, 27-Aug-2018.) |
⊢ (𝑁 ∈ (ℤ≥‘𝑀) → (0 ≤ 𝑀 → 𝑁 ∈ ℕ0)) | ||
Theorem | nltle2tri 44693 | Negated extended trichotomy law for 'less than' and 'less than or equal to'. (Contributed by AV, 18-Jul-2020.) |
⊢ ((𝐴 ∈ ℝ* ∧ 𝐵 ∈ ℝ* ∧ 𝐶 ∈ ℝ*) → ¬ (𝐴 < 𝐵 ∧ 𝐵 ≤ 𝐶 ∧ 𝐶 ≤ 𝐴)) | ||
Theorem | ssfz12 44694 | Subset relationship for finite sets of sequential integers. (Contributed by Alexander van der Vekens, 16-Mar-2018.) |
⊢ ((𝐾 ∈ ℤ ∧ 𝐿 ∈ ℤ ∧ 𝐾 ≤ 𝐿) → ((𝐾...𝐿) ⊆ (𝑀...𝑁) → (𝑀 ≤ 𝐾 ∧ 𝐿 ≤ 𝑁))) | ||
Theorem | elfz2z 44695 | Membership of an integer in a finite set of sequential integers starting at 0. (Contributed by Alexander van der Vekens, 25-May-2018.) |
⊢ ((𝐾 ∈ ℤ ∧ 𝑁 ∈ ℤ) → (𝐾 ∈ (0...𝑁) ↔ (0 ≤ 𝐾 ∧ 𝐾 ≤ 𝑁))) | ||
Theorem | 2elfz3nn0 44696 | If there are two elements in a finite set of sequential integers starting at 0, these two elements as well as the upper bound are nonnegative integers. (Contributed by Alexander van der Vekens, 7-Apr-2018.) |
⊢ ((𝐴 ∈ (0...𝑁) ∧ 𝐵 ∈ (0...𝑁)) → (𝐴 ∈ ℕ0 ∧ 𝐵 ∈ ℕ0 ∧ 𝑁 ∈ ℕ0)) | ||
Theorem | fz0addcom 44697 | The addition of two members of a finite set of sequential integers starting at 0 is commutative. (Contributed by Alexander van der Vekens, 22-May-2018.) (Revised by Alexander van der Vekens, 9-Jun-2018.) |
⊢ ((𝐴 ∈ (0...𝑁) ∧ 𝐵 ∈ (0...𝑁)) → (𝐴 + 𝐵) = (𝐵 + 𝐴)) | ||
Theorem | 2elfz2melfz 44698 | If the sum of two integers of a 0-based finite set of sequential integers is greater than the upper bound, the difference between one of the integers and the difference between the upper bound and the other integer is in the 0-based finite set of sequential integers with the first integer as upper bound. (Contributed by Alexander van der Vekens, 7-Apr-2018.) (Revised by Alexander van der Vekens, 31-May-2018.) |
⊢ ((𝐴 ∈ (0...𝑁) ∧ 𝐵 ∈ (0...𝑁)) → (𝑁 < (𝐴 + 𝐵) → (𝐵 − (𝑁 − 𝐴)) ∈ (0...𝐴))) | ||
Theorem | fz0addge0 44699 | The sum of two integers in 0-based finite sets of sequential integers is greater than or equal to zero. (Contributed by Alexander van der Vekens, 8-Jun-2018.) |
⊢ ((𝐴 ∈ (0...𝑀) ∧ 𝐵 ∈ (0...𝑁)) → 0 ≤ (𝐴 + 𝐵)) | ||
Theorem | elfzlble 44700 | Membership of an integer in a finite set of sequential integers with the integer as upper bound and a lower bound less than or equal to the integer. (Contributed by AV, 21-Oct-2018.) |
⊢ ((𝑁 ∈ ℤ ∧ 𝑀 ∈ ℕ0) → 𝑁 ∈ ((𝑁 − 𝑀)...𝑁)) |
< Previous Next > |
Copyright terms: Public domain | < Previous Next > |