Height one multiple zeta values
DefinitionHeight one multiple zeta function#✓fully proved
For a complex number \(s\) and an integer \(j \ge 0\), the height one multiple zeta function is
where \(T\) runs over the \(j\)-element subsets of \(\{1, \dots, n\}\). Writing \(n_1 = n + 1\) and \(T = \{n_2, \dots, n_{j+1}\}\), this is \(\zeta(s, \{1\}^j) = \sum_{n_1 > \cdots > n_{j+1} \ge 1} n_1^{-s} (n_2 \cdots n_{j+1})^{-1}\), which converges for \(\operatorname{Re} s > 1\).
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Section 1.
Used by 8
TheoremHeight one multiple zeta function of depth one#✓fully proved
For every complex number \(s\), the height one multiple zeta function \(\zeta(s, \{1\}^j)\) at \(j = 0\) is the Dirichlet series of the Riemann zeta function: \(\zeta(s, \{1\}^0) = \sum_{n \ge 0} (n + 1)^{-s}\).
public theorem heightOneMultipleZeta_zero (s : ℂ) :
heightOneMultipleZeta s 0 = ∑' n : ℕ, ((n : ℂ) + 1) ^ (-s)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Section 1.
Uses 1
DefinitionMultiple Hurwitz zeta function#✓fully proved
For complex numbers \(r\) and \(s\), the multiple Hurwitz zeta function of order \(r\), with Hurwitz parameter \(a = 1\), is the Dirichlet series
where \(\binom{m + r - 1}{m} = r (r + 1) \cdots (r + m - 1) / m!\). The series converges for \(\operatorname{Re} r < \operatorname{Re} s\).
public noncomputable def multipleHurwitzZeta (r s : ℂ) : ℂ :=
∑' m : ℕ, (ascPochhammer ℂ m).eval r / (m.factorial : ℂ) * ((m : ℂ) + 1) ^ (-s)
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Section 2, Equation (2.2).
TheoremMultiple Hurwitz zeta function of order zero#✓fully proved
For every complex number \(s\), the multiple Hurwitz zeta function of order \(0\) is identically one: \(\zeta_0(s) = 1\).
public theorem multipleHurwitzZeta_zero (s : ℂ) : multipleHurwitzZeta 0 s = 1
public noncomputable def multipleHurwitzZeta (r s : ℂ) : ℂ :=
∑' m : ℕ, (ascPochhammer ℂ m).eval r / (m.factorial : ℂ) * ((m : ℂ) + 1) ^ (-s)
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:59
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Section 2, Equation (2.2).
Uses 1
TheoremExpansion of the multiple Hurwitz zeta function in height one multiple zeta values#●ready to prove
Let \(r\) and \(s\) be complex numbers with \(\operatorname{Re} s > 1\), \(|r - 1| < |s - 1|\) and \(\operatorname{Re} r < \operatorname{Re} s\). Write \(\zeta(s, \{1\}^j)\) for the height one multiple zeta function and \(\zeta_r(s)\) for the multiple Hurwitz zeta function of order \(r\). Then the series \(\sum_{j \ge 0} \zeta(s, \{1\}^j) \, (r - 1)^j\) is summable with sum \(\zeta_r(s)\).
public theorem_wanted hasSum_heightOneMultipleZeta_mul_pow {s r : ℂ} (hs : 1 < s.re)
(hr : ‖r - 1‖ < ‖s - 1‖) (hrs : r.re < s.re) :
HasSum (fun j : ℕ => heightOneMultipleZeta s j * (r - 1) ^ j) (multipleHurwitzZeta r s)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
public noncomputable def multipleHurwitzZeta (r s : ℂ) : ℂ :=
∑' m : ℕ, (ascPochhammer ℂ m).eval r / (m.factorial : ℂ) * ((m : ℂ) + 1) ^ (-s)
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:59
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Theorem 1.
Uses 2
TheoremHeight one multiple zeta series weighted by powers of a natural number#●ready to prove
Let \(n \ge 0\) be an integer. Write \(\zeta(s, \{1\}^j)\) for the height one multiple zeta function, \(\zeta\) for the Riemann zeta function, \(c(n, m)\) for the unsigned Stirling numbers of the first kind and \(B_j^{(k)}(x)\) for the poly-Bernoulli polynomials. For every complex \(s\) with \(\operatorname{Re} s > n + 1\), the series \(\sum_{j \ge 0} \zeta(s, \{1\}^j) \, n^j\) (with \(0^0 = 1\)) is summable with sum \(\frac{1}{n!} \sum_{m=0}^{n} c(n, m) \, \zeta(s - m)\). For every complex \(s\) with \(\operatorname{Re} s > 1\), the series \(\sum_{j \ge 0} B_j^{(-n)}(-1) \, \zeta(s + n, \{1\}^j)\) is summable with sum \(\zeta(s)\).
The source states the second identity with \(C_j^{(-n)} = B_j^{(-n)}(1)\), where its poly-Bernoulli polynomials have the generating function with \(e^{-xt}\); with the factor \(e^{xt}\) used here this coefficient is \(B_j^{(-n)}(-1)\).
public theorem_wanted hasSum_heightOneMultipleZeta_mul_nat_pow (n : ℕ) :
(∀ s : ℂ, (n : ℝ) + 1 < s.re →
HasSum (fun j : ℕ => heightOneMultipleZeta s j * (n : ℂ) ^ j)
(1 / (n.factorial : ℂ) *
∑ m ∈ Finset.range (n + 1), (n.stirlingFirst m : ℂ) * riemannZeta (s - m))) ∧
∀ s : ℂ, 1 < s.re →
HasSum (fun j : ℕ => (polyBernoulliPoly j (-(n : ℤ)) (-1) : ℂ) *
heightOneMultipleZeta (s + n) j) (riemannZeta s)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
public def polyBernoulliPoly (n : ℕ) (p : ℤ) (x : ℚ) : ℚ :=
∑ i ∈ Finset.range (n + 1),
(n.choose i : ℚ) * polyBernoulli i p * x ^ (n - i)
WantedExt/NumberTheory/BernoulliStirlingHarmonicWanted.lean:35
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Corollary 1.
Uses 2
TheoremHeight one multiple zeta series weighted by powers of a negative integer#●ready to prove
Let \(r \ge 1\) be an integer and let \(\zeta(s, \{1\}^j)\) denote the height one multiple zeta function. For every complex number \(s\) with \(\operatorname{Re} s > r + 1\),
For every integer \(k > r + 1\),
where \(H_{r,k-1} = \sum_{i=1}^{r} (-1)^{i-1} \binom{r}{i} i^{-(k-1)}\) is the rational Roman harmonic number (Equation (2.18)).
public theorem_wanted tsum_heightOneMultipleZeta_mul_neg_nat_pow
(r : ℕ) (hr : 1 ≤ r) :
(∀ s : ℂ, (r : ℝ) + 1 < s.re →
∑' j : ℕ, heightOneMultipleZeta s j * (-(r : ℂ)) ^ j =
∑ m ∈ Finset.range r, ((r - 1).choose m : ℂ) * (-1) ^ m * ((m : ℂ) + 1) ^ (-s)) ∧
∀ k : ℕ, r + 1 < k →
∑' j : ℕ, heightOneMultipleZeta k j * (-(r : ℂ)) ^ j =
(((∑ i ∈ Finset.Icc 1 r, (-1 : ℚ) ^ (i - 1) * (r.choose i : ℚ) / (i : ℚ) ^ (k - 1)) /
r : ℚ) : ℂ)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Corollary 2.
Uses 1
TheoremGeneralized Goldbach series for height one multiple zeta values#●ready to prove
Let \(k \ge 1\) be an integer. Write \(\zeta(s, \{1\}^j)\) for the height one multiple zeta function, \(\zeta\) for the Riemann zeta function, \(c(k, m)\) for the unsigned Stirling numbers of the first kind, \(H_k = \sum_{i=1}^{k} 1/i\) for the \(k\)-th harmonic number and \(H_{k,k} = \sum_{i=1}^{k} (-1)^{i-1} \binom{k}{i} i^{-k}\) for the Roman harmonic number. Then the series \(\sum_{j \ge 0} \bigl(k^j \zeta(k + 1, \{1\}^j) - \frac{1}{k! \, k}\bigr)\) is summable with sum \(\frac{1}{k!} \bigl(\sum_{m=0}^{k-1} c(k, m) \, \zeta(k + 1 - m) + H_k\bigr)\), and the series \(\sum_{j \ge 0} (-1)^j \bigl(k^j \zeta(k + 1, \{1\}^j) - \frac{1}{k! \, k}\bigr)\) is summable with sum \(\frac{H_{k,k}}{k} - \frac{1}{2 \, k! \, k}\).
public theorem_wanted hasSum_heightOneMultipleZeta_goldbach (k : ℕ) (hk : 1 ≤ k) :
HasSum (fun j : ℕ => (k : ℂ) ^ j * heightOneMultipleZeta ((k : ℂ) + 1) j -
1 / ((k.factorial : ℂ) * k))
(1 / (k.factorial : ℂ) *
(∑ m ∈ Finset.range k, (k.stirlingFirst m : ℂ) * riemannZeta ((k : ℂ) + 1 - m) +
(harmonic k : ℂ))) ∧
HasSum (fun j : ℕ => (-1) ^ j * ((k : ℂ) ^ j * heightOneMultipleZeta ((k : ℂ) + 1) j -
1 / ((k.factorial : ℂ) * k)))
(((∑ i ∈ Finset.Icc 1 k, (-1 : ℚ) ^ (i - 1) * (k.choose i : ℚ) / (i : ℚ) ^ k) / k -
1 / (2 * (k.factorial : ℚ) * k) : ℚ) : ℂ)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Theorem 2.
Uses 1
TheoremExpansion of the hyperharmonic zeta function in height one multiple zeta values#●ready to prove
Let \(r\) and \(s\) be complex numbers with \(\operatorname{Re} s > 0\), \(|r| < |s|\) and \(\operatorname{Re} r < \operatorname{Re} s\). Write \(\zeta(s, \{1\}^j)\) for the height one multiple zeta function, \(H_m^{[r]}(k)\) for the generalized hyperharmonic numbers and \(H_m^{[r]} = H_m^{[r]}(1)\); equivalently, \(H_m^{[r]}\) is the derivative of \(\binom{m + r - 1}{m}\) with respect to \(r\). Then the series \(\sum_{j \ge 0} (j + 1) \, \zeta(s + 1, \{1\}^j) \, r^j\) is summable with sum the hyperharmonic zeta function \(\sigma(r, s) = \sum_{m \ge 1} H_m^{[r]} m^{-s}\).
public theorem_wanted hasSum_succ_mul_heightOneMultipleZeta_mul_pow {r s : ℂ} (hs : 0 < s.re)
(hr : ‖r‖ < ‖s‖) (hrs : r.re < s.re) :
HasSum (fun j : ℕ => ((j : ℂ) + 1) * heightOneMultipleZeta (s + 1) j * r ^ j)
(∑' m : ℕ, generalizedHyperharmonic (m + 1) 1 r * ((m : ℂ) + 1) ^ (-s))
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
noncomputable def generalizedHyperharmonic (n k : ℕ) (r : ℂ) : ℂ :=
if k ≤ n then
(∑ s ∈ (Finset.range n).powersetCard (n - k), ∏ j ∈ s, (r + j)) / n.factorial
else 0
WantedExt/NumberTheory/HyperharmonicEulerSumsWanted.lean:34
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Theorem 3, second equality.
Uses 2
TheoremShifted alternating series of height one multiple zeta values#●ready to prove
Let \(k \ge -1\) and \(p \ge 0\) be integers, write \(\zeta(s, \{1\}^p)\) for the height one multiple zeta function and \(\zeta_r(s)\) for the multiple Hurwitz zeta function of order \(r\). Then the partial sums of the series \(\nu_{k,p} = \sum_{n \ge 2} \frac{(-1)^n \, \zeta(n, \{1\}^p)}{n + k}\) converge to \(\int_0^1 (1 - r)^{k+1} \, \zeta_r(p + 2) \, dr\).
public theorem_wanted tendsto_alternating_heightOneMultipleZeta_integral (k : ℤ) (hk : -1 ≤ k)
(p : ℕ) :
Filter.Tendsto
(fun N : ℕ => ∑ n ∈ Finset.Ico 2 N,
(-1 : ℂ) ^ n / ((n : ℂ) + k) * heightOneMultipleZeta n p)
Filter.atTop
(nhds (∫ r in (0 : ℝ)..1,
(1 - (r : ℂ)) ^ (k + 1).toNat * multipleHurwitzZeta r ((p : ℂ) + 2)))
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
public noncomputable def multipleHurwitzZeta (r s : ℂ) : ℂ :=
∑' m : ℕ, (ascPochhammer ℂ m).eval r / (m.factorial : ℂ) * ((m : ℂ) + 1) ^ (-s)
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:59
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Theorem 4.
Uses 2
TheoremLinear Laurent coefficient of the height one multiple zeta function#●ready to prove
Let \(j \ge 1\) be an integer, write \(\zeta(s, \{1\}^{j-1})\) for the height one multiple zeta function and \(\zeta_r(s)\) for the multiple Hurwitz zeta function of order \(r\). There are complex numbers \(c_0, \dots, c_{j-1}\) and a function \(F\) analytic at \(s = 1\) such that \(\zeta(s, \{1\}^{j-1}) = \sum_{i=0}^{j-1} c_i (s - 1)^{-(i+1)} + F(s)\) for all \(s\) near \(1\) with \(\operatorname{Re} s > 1\), and
where \(a_{j+1}\) is the coefficient of \(s^{j+1}\) in the Taylor expansion of \(1/\Gamma(s + 1)\) at \(s = 0\).
The number \(-F'(1)\) generalizes the first Stieltjes constant \(\gamma_1\): for \(j = 1\) and \(\operatorname{Re} s > 1\), \(\zeta(s, \{1\}^0)\) equals the Riemann zeta function \(\zeta(s)\), whose Laurent expansion at \(s = 1\) is \(\zeta(s) = \frac{1}{s - 1} + \gamma - \gamma_1 (s - 1) + \cdots\) with Euler's constant \(\gamma\), so \(-F'(1) = \gamma_1\).
public theorem_wanted heightOneMultipleZeta_laurent_linear_coeff (j : ℕ) (hj : 1 ≤ j) :
∃ c : ℕ → ℂ, ∃ F : ℂ → ℂ, AnalyticAt ℂ F 1 ∧
(∀ᶠ s in nhdsWithin (1 : ℂ) {s | 1 < s.re},
heightOneMultipleZeta s (j - 1) =
∑ i ∈ Finset.range j, c i / (s - 1) ^ (i + 1) + F s) ∧
-deriv F 1 =
(∫ r in (0 : ℝ)..1, ((1 - (r : ℂ)) * multipleHurwitzZeta r ((j : ℂ) + 2) -
multipleHurwitzZeta r ((j : ℂ) + 1))) -
iteratedDeriv (j + 1) (fun s : ℂ => (Complex.Gamma (s + 1))⁻¹) 0 /
((j + 1).factorial : ℂ)
public noncomputable def heightOneMultipleZeta (s : ℂ) (j : ℕ) : ℂ :=
∑' n : ℕ, ((n : ℂ) + 1) ^ (-s) *
∑ T ∈ (Finset.Icc 1 n).powersetCard j, ∏ m ∈ T, (1 / (m : ℂ))
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:42
public noncomputable def multipleHurwitzZeta (r s : ℂ) : ℂ :=
∑' m : ℕ, (ascPochhammer ℂ m).eval r / (m.factorial : ℂ) * ((m : ℂ) + 1) ^ (-s)
WantedExt/NumberTheory/HeightOneMultipleZetaSeriesWanted.lean:59
Sources
- Paul Thomas Young, Series of Height One Multiple Zeta Functions, INTEGERS 24 (2024), Article A43, Corollary 3.
Uses 2
Loading dependency explorer…
The dependency explorer needs JavaScript and its graph data.