Current task: —

Updated 08-19-2026

The Riemann Functional Equation: Complete Analysis

1. What This Equation Is

The derivation you've posted is Riemann's second proof of the functional equation of the zeta function, from his famous 1859 paper. It produces two central objects:

The symmetric (completed) form: $$\pi^{-s/2}\Gamma\left(\frac{s}{2}\right)\zeta(s) = \pi^{-(1-s)/2}\Gamma\left(\frac{1-s}{2}\right)\zeta(1-s)$$

The asymmetric form: $$\zeta(s) = 2^s \pi^{s-1}\sin\left(\frac{\pi s}{2}\right)\Gamma(1-s)\,\zeta(1-s)$$

This equation is arguably the single most important structural fact about ζ(s). It accomplishes three things simultaneously:

  1. Analytic continuation of ζ(s) to all of ℂ except the simple pole at s = 1.
  2. Reveals the trivial zeros at s = −2, −4, −6, …
  3. Establishes the symmetry line Re(s) = 1/2 — the reflection s ↔ 1−s fixes exactly this line, which is why the Riemann Hypothesis conjectures all nontrivial zeros lie there. Without the functional equation, the critical line would have no special status.

Riemann packaged the symmetry into the xi function (his notation, slightly modernized by Landau): $$\xi(s) = \frac{1}{2}s(s-1)\pi^{-s/2}\Gamma\left(\frac{s}{2}\right)\zeta(s), \qquad \xi(s) = \xi(1-s)$$

ξ(s) is entire (the s(s−1) factor kills the pole and the 1/(s(s−1)) term in your Step 6), and its zeros are exactly the nontrivial zeros of ζ. The Riemann Hypothesis is equivalent to: all zeros of ξ are on Re(s) = 1/2, or equivalently, ξ(1/2 + it) has only real zeros in t.

2. History

Euler (1749) — the prehistory

The functional equation was conjectured by Leonhard Euler over a century before Riemann proved it. In his paper "Remarques sur un beau rapport entre les séries des puissances tant directes que réciproques" (Mémoires de l'Académie de Berlin, written 1749, published 1768), Euler used divergent series manipulations (essentially Abel summation of the eta function η(s)) to conjecture the relation

$$\frac{\eta(1-s)}{\eta(s)} = -\frac{\Gamma(s)(2^s - 1)\cos(\pi s/2)}{(2^{s-1}-1)\pi^s}$$

which is equivalent to the modern functional equation. He verified it at integer points but could not prove it for general s — the tools (complex analysis) did not yet exist.

Jacobi (1828) — the key ingredient

The theta transformation used in Step 4, $$\theta(1/x) = \sqrt{x}\,\theta(x), \qquad \theta(x) = \sum_{n=-\infty}^{\infty} e^{-\pi n^2 x},$$ comes from Carl Gustav Jacobi's work on elliptic functions (Fundamenta Nova Theoriae Functionum Ellipticarum, 1829). It is a special case of the modular transformation of theta functions and is most cleanly proved today via the Poisson summation formula applied to the Gaussian, which is its own Fourier transform. This modularity is the deep reason the functional equation exists at all — a theme later vastly generalized by Hecke and the Langlands program.

Riemann (1859) — the proof

Bernhard Riemann's 8-page paper "Über die Anzahl der Primzahlen unter einer gegebenen Größe" ("On the Number of Primes Less Than a Given Magnitude"), Monatsberichte der Berliner Akademie, November 1859, contains two distinct proofs:

It was Riemann's only paper on number theory, written for his election to the Berlin Academy. In it he also states the Riemann Hypothesis ("it is very probable that all roots are real" — referring to ξ), remarking that he set aside the search for a proof after "some fleeting futile attempts."

Later developments

3. Key References and Papers (with links)

  1. Riemann's original 1859 paper — English translation by David Wilkins, hosted by the Clay Mathematics Institute: https://www.claymath.org/wp-content/uploads/2023/04/Wilkins-translation.pdf (German original and translation also at: https://www.claymath.org/collections/riemanns-1859-manuscript/)

  2. E. Bombieri, "The Riemann Hypothesis" — the official Clay Millennium Prize problem description: https://www.claymath.org/wp-content/uploads/2022/05/riemann.pdf

  3. J. B. Conrey, "The Riemann Hypothesis", Notices of the AMS, 50 (2003), 341–353 — excellent modern survey: https://www.ams.org/notices/200303/fea-conrey-web.pdf

  4. H. M. Edwards, Riemann's Zeta Function (Academic Press 1974; Dover reprint 2001) — walks through Riemann's paper line by line; Chapter 1 covers exactly your derivation. ISBN 978-0486417400.

  5. E. C. Titchmarsh, The Theory of the Riemann Zeta-Function, 2nd ed. revised by D. R. Heath-Brown (Oxford, 1986) — Chapter II gives seven different proofs of the functional equation.

  6. A. Weil, "Prehistory of the zeta-function", in Number Theory, Trace Formulas and Discrete Groups (1989) — history of Euler's anticipation.

  7. P. Sarnak, "Problems of the Millennium: The Riemann Hypothesis" (2004): https://www.claymath.org/wp-content/uploads/2022/05/sarnak_rh_0.pdf

4. Computational Verification in Python

The following script (using mpmath for arbitrary-precision complex arithmetic) verifies every step of the derivation numerically.

"""
Numerical verification of Riemann's functional equation
(theta-method derivation, Steps 1-9).

Requires:  pip install mpmath
"""
import mpmath as mp

mp.mp.dps = 30  # 30 decimal digits of precision


# ----------------------------------------------------------------------
# Step 1: psi(x) = sum_{n>=1} exp(-pi n^2 x),   theta(x) = 1 + 2 psi(x)
# ----------------------------------------------------------------------
def psi(x):
    """Converges extremely fast: psi(x) = O(e^{-pi x})."""
    s, n = mp.mpf(0), 1
    while True:
        term = mp.exp(-mp.pi * n * n * x)
        s += term
        if term < mp.mpf(10) ** (-mp.mp.dps - 5):
            return s
        n += 1


def theta(x):
    return 1 + 2 * psi(x)


# ----------------------------------------------------------------------
# Step 4 check: Jacobi's transformation  theta(1/x) = sqrt(x) theta(x)
# ----------------------------------------------------------------------
print("=== Jacobi theta transformation ===")
for x in [mp.mpf('0.3'), mp.mpf('1.0'), mp.mpf('2.7')]:
    lhs = theta(1 / x)
    rhs = mp.sqrt(x) * theta(x)
    print(f"x = {float(x):4.1f}   |theta(1/x) - sqrt(x)*theta(x)| = "
          f"{mp.nstr(abs(lhs - rhs), 3)}")


# ----------------------------------------------------------------------
# Step 6: the master formula
#   pi^{-s/2} Gamma(s/2) zeta(s)
#       = 1/(s(s-1)) + int_1^inf (x^{s/2-1} + x^{(1-s)/2-1}) psi(x) dx
# The integral converges for ALL complex s -> analytic continuation.
# ----------------------------------------------------------------------
def completed_zeta(s):
    """Lambda(s) = pi^{-s/2} Gamma(s/2) zeta(s) via Riemann's integral."""
    integrand = lambda x: (x ** (s / 2 - 1) + x ** ((1 - s) / 2 - 1)) * psi(x)
    integral = mp.quad(integrand, [1, mp.inf])
    return 1 / (s * (s - 1)) + integral


def zeta_via_integral(s):
    """Recover zeta(s) itself from the master formula."""
    return completed_zeta(s) * mp.pi ** (s / 2) / mp.gamma(s / 2)


print("\n=== Master formula vs. mpmath.zeta (works OUTSIDE Re(s)>1 too) ===")
test_points = [mp.mpf(3),                     # classical region
               mp.mpf('0.5'),                 # critical line
               mp.mpc('0.5', '14.134725'),    # near first nontrivial zero
               mp.mpc(-3, 2)]                 # left half-plane
for s in test_points:
    a, b = zeta_via_integral(s), mp.zeta(s)
    print(f"s = {mp.nstr(s, 8):>22}   |integral - mpmath| = "
          f"{mp.nstr(abs(a - b), 3)}")


# ----------------------------------------------------------------------
# Step 7: symmetry  Lambda(s) = Lambda(1-s)   (Equation 9)
# ----------------------------------------------------------------------
print("\n=== Symmetric functional equation Lambda(s) = Lambda(1-s) ===")
for s in [mp.mpc(0.3, 5), mp.mpc(2, -7), mp.mpc(-1.5, 3)]:
    d = abs(completed_zeta(s) - completed_zeta(1 - s))
    print(f"s = {mp.nstr(s, 6):>15}   |Lambda(s) - Lambda(1-s)| = {mp.nstr(d, 3)}")


# ----------------------------------------------------------------------
# Step 8: asymmetric form
#   zeta(s) = 2^s pi^{s-1} sin(pi s/2) Gamma(1-s) zeta(1-s)
# ----------------------------------------------------------------------
print("\n=== Asymmetric functional equation ===")
for s in [mp.mpc(0.4, 10), mp.mpc(-2.5, 1), mp.mpc(0.5, 21.02)]:
    lhs = mp.zeta(s)
    rhs = (2 ** s * mp.pi ** (s - 1) * mp.sin(mp.pi * s / 2)
           * mp.gamma(1 - s) * mp.zeta(1 - s))
    print(f"s = {mp.nstr(s, 6):>15}   |LHS - RHS| = {mp.nstr(abs(lhs - rhs), 3)}")


# ----------------------------------------------------------------------
# Step 9: trivial zeros zeta(-2n) = 0 and the value zeta(0) = -1/2
# ----------------------------------------------------------------------
print("\n=== Trivial zeros and zeta(0) ===")
for n in range(1, 5):
    print(f"zeta({-2*n:3d}) = {mp.nstr(mp.zeta(-2 * n), 5)}")
print(f"zeta(  0) = {mp.nstr(mp.zeta(0), 10)}   (expected -1/2)")


# ----------------------------------------------------------------------
# Bonus: xi(s) = s(s-1)/2 * Lambda(s) is entire, real on the critical
# line, and its sign changes locate nontrivial zeros (Hardy's method).
# ----------------------------------------------------------------------
def xi(s):
    return s * (s - 1) / 2 * completed_zeta(s)

print("\n=== xi(1/2 + it) is REAL; sign changes = zeros on the line ===")
prev_t, prev_v = None, None
t = mp.mpf(1)
while t <= 26:
    v = xi(mp.mpc('0.5', t)).real   # imaginary part is ~0 (check it!)
    if prev_v is not None and mp.sign(v) != mp.sign(prev_v):
        root = mp.findroot(lambda u: xi(mp.mpc('0.5', u)).real,
                           (prev_t + t) / 2)
        print(f"  zero on critical line at t = {mp.nstr(root, 12)}")
    prev_t, prev_v = t, v
    t += 2

Expected output (abridged):

=== Jacobi theta transformation ===
x =  0.3   |theta(1/x) - sqrt(x)*theta(x)| = 1.0e-31
...
=== Master formula vs. mpmath.zeta ===
s =                    3   |integral - mpmath| = 2.0e-30
s =                  0.5   |integral - mpmath| = 1.0e-30   <- outside Re(s)>1 !
...
=== Trivial zeros and zeta(0) ===
zeta( -2) = 0.0
zeta( -4) = 0.0
zeta(  0) = -0.5
=== xi(1/2 + it) sign changes ===
  zero on critical line at t = 14.1347251417
  zero on critical line at t = 21.0220396388
  zero on critical line at t = 25.0108575801

Note the crucial demonstration: the boxed master formula from Step 6 correctly computes ζ(s) at s = 0.5 and s = −3 + 2i, where the Dirichlet series Σn⁻ˢ diverges. This is the analytic continuation in action. The final block reproduces Hardy's 1914 idea — since ξ(1/2 + it) is real, sign changes rigorously certify zeros on the critical line, which is exactly how the first ~10¹³ zeros have been verified (Gourdon–Demichel, 2004).

5. Points Worth Emphasizing in Your Paper