Large Zsigmondy primes
DefinitionLarge Zsigmondy prime#✓fully proved
Let \(a\), \(b\), \(n\) and \(p\) be natural numbers. Then \(p\) is a large Zsigmondy prime for \((a, b, n)\) when \(p\) is a primitive prime divisor of the term \(a^n - b^n\) of the integer sequence \((a^m - b^m)_{m \ge 0}\), that is, \(n \ge 1\), \(p\) is prime, \(p\) divides \(a^n - b^n\) and \(p\) divides no \(a^m - b^m\) with \(1 \le m < n\), and moreover \(p^2\) divides \(a^n - b^n\) or \(p > n + 1\).
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
def IsPrimitivePrimeDivisor (a : ℕ → ℤ) (m u : ℕ) : Prop :=
Nat.Prime u ∧ IsPrimitiveDivisor a m u
MathlibExt/NumberTheory/PrimitiveDivisor.lean:41
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Section 1 (definition after Theorem 1).
Uses 1 · Used by 5
Theorem\(7\) is a large Zsigmondy prime for \((2, 1, 3)\)#✓fully proved
The prime \(7\) is a large Zsigmondy prime for \((a, b, n) = (2, 1, 3)\): it divides \(2^3 - 1 = 7\), divides neither \(2^1 - 1\) nor \(2^2 - 1\), and \(7 > 3 + 1\).
theorem isLargeZsigmondyPrime_two_one_three_seven : IsLargeZsigmondyPrime 2 1 3 7
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
WantedExt/NumberTheory/LargeZsigmondyPrimesWanted.lean:44
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Section 1 (definition after Theorem 1).
Uses 3
TheoremSquare-root lower bound for Euler's totient#●ready to prove
For every positive integer \(n\), \(\varphi(n) \ge \sqrt{n}/2\), where \(\varphi\) is Euler's totient function.
theorem_wanted sqrt_div_two_le_totient (n : ℕ) (hn : 0 < n) :
Real.sqrt n / 2 ≤ (Nat.totient n : ℝ)
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 1 (quoted from reference [2]).
TheoremHomogenized cyclotomic values at \(p^\beta k\)#●ready to prove
Let \(p\) be a prime, let \(\beta \ge 1\) and \(k \ge 1\) be integers with \(p \nmid k\), and let \(a, b\) be integers; write \(\Phi_m(a, b)\) for the homogenized cyclotomic value. Then \(\Phi_{p^\beta k}(a, b) = \Phi_{pk}(a^{p^{\beta - 1}}, b^{p^{\beta - 1}})\) and \(\Phi_{p^\beta k}(a, b) \cdot \Phi_k(a^{p^{\beta - 1}}, b^{p^{\beta - 1}}) = \Phi_k(a^{p^\beta}, b^{p^\beta})\).
The paper states the second identity as a quotient; here it is cleared of its denominator.
theorem_wanted homogenizedCyclotomic_prime_pow_mul {p β k : ℕ} (hp : p.Prime) (hβ : 0 < β)
(hk : 0 < k) (hpk : ¬ p ∣ k) (a b : ℤ) :
homogenizedCyclotomic (p ^ β * k) a b =
homogenizedCyclotomic (p * k) (a ^ p ^ (β - 1)) (b ^ p ^ (β - 1)) ∧
homogenizedCyclotomic (p ^ β * k) a b *
homogenizedCyclotomic k (a ^ p ^ (β - 1)) (b ^ p ^ (β - 1)) =
homogenizedCyclotomic k (a ^ p ^ β) (b ^ p ^ β)
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Corollary 1.
Uses 1
TheoremFactorization of \(a^n - b^n\) into homogenized cyclotomic values#●ready to prove
For every positive integer \(n\) and all integers \(a\) and \(b\),
where the product runs over the positive divisors \(d\) of \(n\) and \(\Phi_d(a, b)\) is the homogenized cyclotomic value.
theorem_wanted prod_divisors_homogenizedCyclotomic {n : ℕ} (hn : 0 < n) (a b : ℤ) :
∏ d ∈ n.divisors, homogenizedCyclotomic d a b = a ^ n - b ^ n
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 3 (i) (quoted from reference [6]).
Uses 1
TheoremBounds for homogenized cyclotomic values#●ready to prove
Let \(a > b\) be positive integers and let \(n \ge 1\). Then the homogenized cyclotomic value \(\Phi_n(a, b)\) satisfies \((a - b)^{\varphi(n)} \le \Phi_n(a, b) \le (a + b)^{\varphi(n)}\), where \(\varphi\) is Euler's totient function.
theorem_wanted homogenizedCyclotomic_mem_Icc {a b n : ℕ} (hb : 0 < b) (hba : b < a)
(hn : 0 < n) :
((a : ℤ) - b) ^ Nat.totient n ≤ homogenizedCyclotomic n a b ∧
homogenizedCyclotomic n a b ≤ ((a : ℤ) + b) ^ Nat.totient n
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 3 (ii) (quoted from reference [6]).
Uses 1
TheoremHomogenized cyclotomic values at \(pn\)#●ready to prove
Let \(p\) be a prime, \(n\) a positive integer and \(a, b\) integers, and write \(\Phi_m(a, b)\) for the homogenized cyclotomic value. If \(p \mid n\), then \(\Phi_{pn}(a, b) = \Phi_n(a^p, b^p)\). If \(p \nmid n\), then \(\Phi_{pn}(a, b) \cdot \Phi_n(a, b) = \Phi_n(a^p, b^p)\).
The paper states the case \(p \nmid n\) as a quotient; here it is cleared of its denominator.
theorem_wanted homogenizedCyclotomic_prime_mul {p n : ℕ} (hp : p.Prime) (hn : 0 < n)
(a b : ℤ) :
(p ∣ n → homogenizedCyclotomic (p * n) a b = homogenizedCyclotomic n (a ^ p) (b ^ p)) ∧
(¬ p ∣ n → homogenizedCyclotomic (p * n) a b * homogenizedCyclotomic n a b =
homogenizedCyclotomic n (a ^ p) (b ^ p))
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 3 (iii) (quoted from reference [6]).
Uses 1
Theorem\(p\)-adic valuation of homogenized cyclotomic values at an odd prime#●ready to prove
Let \(a > b\) be positive integers, let \(p\) be an odd prime dividing neither \(a\) nor \(b\), and let \(k\) be the least positive integer such that \(p\) divides \(a^k - b^k\). Write \(v_p\) for the \(p\)-adic valuation and \(\Phi_n(a, b)\) for the homogenized cyclotomic value. Then for every positive integer \(n\): \(v_p(\Phi_n(a, b)) = v_p(a^k - b^k)\) if \(n = k\); \(v_p(\Phi_n(a, b)) = 1\) if \(n = p^\beta k\) for some \(\beta \ge 1\); and \(v_p(\Phi_n(a, b)) = 0\) otherwise.
theorem_wanted padicValInt_homogenizedCyclotomic_odd {a b n p k : ℕ} (hb : 0 < b)
(hba : b < a) (hn : 0 < n) (hp : p.Prime) (hp2 : p ≠ 2) (hpa : ¬ p ∣ a) (hpb : ¬ p ∣ b)
(hk : 0 < k) (hpk : (p : ℤ) ∣ (a : ℤ) ^ k - (b : ℤ) ^ k)
(hkmin : ∀ m, 0 < m → m < k → ¬ (p : ℤ) ∣ (a : ℤ) ^ m - (b : ℤ) ^ m) :
(n = k → padicValInt p (homogenizedCyclotomic n a b) =
padicValInt p ((a : ℤ) ^ k - (b : ℤ) ^ k)) ∧
((∃ β, 1 ≤ β ∧ n = p ^ β * k) → padicValInt p (homogenizedCyclotomic n a b) = 1) ∧
(n ≠ k → (¬ ∃ β, 1 ≤ β ∧ n = p ^ β * k) →
padicValInt p (homogenizedCyclotomic n a b) = 0)
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 3 (iv) (quoted from reference [6]).
Uses 1
Theorem\(2\)-adic valuation of homogenized cyclotomic values#●ready to prove
Let \(a > b\) be odd positive integers. Write \(v_2\) for the \(2\)-adic valuation and \(\Phi_n(a, b)\) for the homogenized cyclotomic value. Then for every positive integer \(n\): \(v_2(\Phi_n(a, b)) = v_2(a - b)\) if \(n = 1\); \(v_2(\Phi_n(a, b)) = v_2(a + b)\) if \(n = 2\); \(v_2(\Phi_n(a, b)) = 1\) if \(n = 2^\beta\) for some \(\beta \ge 2\); and \(v_2(\Phi_n(a, b)) = 0\) if \(n\) is not a power of \(2\).
theorem_wanted padicValInt_two_homogenizedCyclotomic {a b n : ℕ} (hb : 0 < b) (hba : b < a)
(ha : Odd a) (hb' : Odd b) (hn : 0 < n) :
(n = 1 → padicValInt 2 (homogenizedCyclotomic n a b) = padicValInt 2 ((a : ℤ) - b)) ∧
(n = 2 → padicValInt 2 (homogenizedCyclotomic n a b) = padicValInt 2 ((a : ℤ) + b)) ∧
((∃ β, 2 ≤ β ∧ n = 2 ^ β) → padicValInt 2 (homogenizedCyclotomic n a b) = 1) ∧
((¬ ∃ β, n = 2 ^ β) → padicValInt 2 (homogenizedCyclotomic n a b) = 0)
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 3 (v) (quoted from reference [6]).
Uses 1
TheoremPrime divisors of homogenized cyclotomic values#●ready to prove
Let \(a > b\) be coprime positive integers, let \(n \ge 2\), let \(\Phi_n(a, b)\) be the homogenized cyclotomic value, let \(p\) be a prime dividing \(\Phi_n(a, b)\), and let \(k\) be the least positive integer such that \(p\) divides \(a^k - b^k\). Then one of the following holds: (i) \(p = 2\) and \(n = 2^\beta\) for some \(\beta \ge 1\); (ii) \(p \ge 3\), \(n = k\), and \(p\) is a primitive prime divisor of the term \(a^n - b^n\) of the sequence \((a^m - b^m)_{m \ge 0}\); (iii) \(p\) is the greatest prime factor of \(n\), \(p > 2\), \(n = p^\beta k\) for some \(\beta \ge 1\), and the \(p\)-adic valuation of \(\Phi_n(a, b)\) is \(1\).
theorem_wanted prime_dvd_homogenizedCyclotomic_cases {a b n p k : ℕ} (hb : 0 < b)
(hba : b < a) (hab : a.Coprime b) (hn : 2 ≤ n) (hp : p.Prime)
(hpΦ : (p : ℤ) ∣ homogenizedCyclotomic n a b) (hk : 0 < k)
(hpk : (p : ℤ) ∣ (a : ℤ) ^ k - (b : ℤ) ^ k)
(hkmin : ∀ m, 0 < m → m < k → ¬ (p : ℤ) ∣ (a : ℤ) ^ m - (b : ℤ) ^ m) :
(p = 2 ∧ ∃ β, 1 ≤ β ∧ n = 2 ^ β) ∨
(3 ≤ p ∧ n = k ∧
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p) ∨
(p ∈ n.primeFactors ∧ (∀ r ∈ n.primeFactors, r ≤ p) ∧ 2 < p ∧
(∃ β, 1 ≤ β ∧ n = p ^ β * k) ∧ padicValInt p (homogenizedCyclotomic n a b) = 1)
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
def IsPrimitivePrimeDivisor (a : ℕ → ℤ) (m u : ℕ) : Prop :=
Nat.Prime u ∧ IsPrimitiveDivisor a m u
MathlibExt/NumberTheory/PrimitiveDivisor.lean:41
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 4.
Uses 2
TheoremZsigmondy primes divide the homogenized cyclotomic value#●ready to prove
Let \(a\) and \(b\) be distinct coprime positive integers and let \(n \ge 2\). If \(p\) is a primitive prime divisor of the term \(a^n - b^n\) of the integer sequence \((a^m - b^m)_{m \ge 0}\) (a Zsigmondy prime for \((a, b, n)\)), then \(p\) divides the homogenized cyclotomic value \(\Phi_n(a, b)\).
theorem_wanted dvd_homogenizedCyclotomic_of_isPrimitivePrimeDivisor {a b n p : ℕ}
(ha : 0 < a) (hb : 0 < b) (hab : a ≠ b) (hcop : a.Coprime b) (hn : 2 ≤ n)
(hp : IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p) :
(p : ℤ) ∣ homogenizedCyclotomic n a b
def IsPrimitivePrimeDivisor (a : ℕ → ℤ) (m u : ℕ) : Prop :=
Nat.Prime u ∧ IsPrimitiveDivisor a m u
MathlibExt/NumberTheory/PrimitiveDivisor.lean:41
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 5.
Uses 2
TheoremA Zsigmondy prime \(q\) that is not large satisfies \(n = q - 1\)#●ready to prove
Let \(a > b\) be coprime positive integers and let \(n \ge 2\). If \(q\) is a primitive prime divisor of the term \(a^n - b^n\) of the integer sequence \((a^m - b^m)_{m \ge 0}\) and \(q\) is not a large Zsigmondy prime for \((a, b, n)\), then \(n = q - 1\).
theorem_wanted eq_sub_one_of_not_isLargeZsigmondyPrime {a b n q : ℕ} (hb : 0 < b)
(hba : b < a) (hcop : a.Coprime b) (hn : 2 ≤ n)
(hq : IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n q)
(hql : ¬ IsLargeZsigmondyPrime a b n q) :
n = q - 1
def IsPrimitivePrimeDivisor (a : ℕ → ℤ) (m u : ℕ) : Prop :=
Nat.Prime u ∧ IsPrimitiveDivisor a m u
MathlibExt/NumberTheory/PrimitiveDivisor.lean:41
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
WantedExt/NumberTheory/LargeZsigmondyPrimesWanted.lean:44
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 6.
Uses 2
TheoremA size criterion for a large Zsigmondy prime#●ready to prove
Let \(a\) and \(b\) be distinct coprime positive integers and let \(n \ge 3\). If \((n + 1) q < \Phi_n(a, b)\) for every prime factor \(q\) of \(n\), where \(\Phi_n(a, b)\) is the homogenized cyclotomic value, then there is a large Zsigmondy prime for \((a, b, n)\).
Equivalently, the hypothesis asks that \(n + 1\) times the greatest prime factor of \(n\) be less than \(\Phi_n(a, b)\).
theorem_wanted exists_isLargeZsigmondyPrime_of_lt {a b n : ℕ} (ha : 0 < a) (hb : 0 < b)
(hab : a ≠ b) (hcop : a.Coprime b) (hn : 3 ≤ n)
(hlt : ∀ q ∈ n.primeFactors, ((n + 1) * q : ℤ) < homogenizedCyclotomic n a b) :
∃ p, IsLargeZsigmondyPrime a b n p
public noncomputable def homogenizedCyclotomic (n : ℕ) (a b : ℤ) : ℤ :=
∑ i ∈ Finset.range (Nat.totient n + 1),
(cyclotomic n ℤ).coeff i * a ^ i * b ^ (Nat.totient n - i)
MathlibExt/NumberTheory/Cyclotomic/HomogenizedCyclotomic.lean:38
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
WantedExt/NumberTheory/LargeZsigmondyPrimesWanted.lean:44
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 7.
Uses 2
TheoremLarge Zsigmondy primes for exponents \(n > 1\) outside six values#●ready to prove
Let \(a > b\) be coprime positive integers and let \(n > 1\) with \(n \notin \{2, 4, 6, 10, 12, 18\}\). Then there is a large Zsigmondy prime for \((a, b, n)\).
theorem_wanted exists_isLargeZsigmondyPrime_of_not_mem {a b n : ℕ} (hb : 0 < b) (hba : b < a)
(hcop : a.Coprime b) (hn : 1 < n) (hn' : n ∉ ({2, 4, 6, 10, 12, 18} : Finset ℕ)) :
∃ p, IsLargeZsigmondyPrime a b n p
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
WantedExt/NumberTheory/LargeZsigmondyPrimesWanted.lean:44
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Lemma 8.
Uses 1
TheoremAvcı's classification of large Zsigmondy primes#●ready to prove
Let \(a > b\) be coprime positive integers and let \(n > 1\). There is a large Zsigmondy prime for \((a, b, n)\) if and only if none of the following holds: \(n = 2\) and \(a + b = 2^s\) or \(a + b = 3 \cdot 2^s\) for some natural number \(s\); \(n = 4\) and \((a, b) \in \{(2, 1), (3, 1)\}\); \(n = 6\) and \((a, b) \in \{(2, 1), (3, 1), (3, 2), (5, 1), (5, 4)\}\); \(n = 10\) and \((a, b) \in \{(2, 1), (3, 2)\}\); \(n = 12\) and \((a, b) = (2, 1)\); \(n = 18\) and \((a, b) = (2, 1)\).
theorem_wanted exists_isLargeZsigmondyPrime_iff {a b n : ℕ} (hb : 0 < b) (hba : b < a)
(hcop : a.Coprime b) (hn : 1 < n) :
(∃ p, IsLargeZsigmondyPrime a b n p) ↔
¬ ((n = 2 ∧ ∃ s : ℕ, a + b = 2 ^ s ∨ a + b = 3 * 2 ^ s) ∨
(n = 4 ∧ (a, b) ∈ ({(2, 1), (3, 1)} : Finset (ℕ × ℕ))) ∨
(n = 6 ∧ (a, b) ∈ ({(2, 1), (3, 1), (3, 2), (5, 1), (5, 4)} : Finset (ℕ × ℕ))) ∨
(n = 10 ∧ (a, b) ∈ ({(2, 1), (3, 2)} : Finset (ℕ × ℕ))) ∨
(n = 12 ∧ (a, b) = (2, 1)) ∨ (n = 18 ∧ (a, b) = (2, 1)))
def IsLargeZsigmondyPrime (a b n p : ℕ) : Prop :=
IntegerSequence.IsPrimitivePrimeDivisor (fun m => (a : ℤ) ^ m - (b : ℤ) ^ m) n p ∧
((p : ℤ) ^ 2 ∣ (a : ℤ) ^ n - (b : ℤ) ^ n ∨ n + 1 < p)
WantedExt/NumberTheory/LargeZsigmondyPrimesWanted.lean:44
Sources
- Ömer Avcı, Large Zsigmondy Primes, INTEGERS 24 (2024), Article A62, Theorem 1 (the converse is the classification in the proof of Theorem 1).
Uses 1
Loading dependency explorer…
The dependency explorer needs JavaScript and its graph data.