Reconstructs the prime-counting staircase ψ(x) as a superposition of spectral waves — one per Riemann zero — making tangible the duality
prime positions ⟷ spectral geometry {γ_n}.The Riemann–von Mangoldt explicit formula expresses the prime staircase ψ(x) = Σ_{p^k ≤ x} log p as
ψ(x) = x − Σ_n x^{ρ_n}/ρ_n − log(2π) − ½ log(1 − x^{-2})with ρ_n = ½ + i·γ_n the non-trivial zeros. Each oscillatory term
x^{ρ_n}/ρ_n = (√x / |ρ_n|) · exp(i·γ_n·log x − i·arg ρ_n)is a WAVE of angular frequency γ_n in the coordinate log x and amplitude √x/|ρ_n|. Primes appear where they do because their distribution is the Fourier synthesis of these spectral waves: the "coherence flow ordering itself" is literally the superposition of e^{i·γ_n·log x}.
RH states Re(ρ_n) = ½ for EVERY zero. Then every wave has amplitude √x/|ρ_n| — uniformly bounded — and the staircase stays coherent (|ψ(x) − x| = O(√x log²x)). A single off-critical-line zero at Re = ½ + δ would inject a wave of amplitude x^{½+δ} that grows faster and DESTROYS the bound. So RH ⟺ "all spectral waves equally bounded" ⟺ the spectral geometry is maximally coherent.
This REPRODUCES the classical explicit formula (Riemann 1859, von Mangoldt 1895); it is NOT a TNFR discovery. TNFR's contribution is to realise the prime-side spectrum as a self-adjoint prime-ladder Hamiltonian (P14, spectrum {k·log p}) and verify the identity to machine precision (P15). The "spectral geometry" is the geometry of an operator on the critical line — NOT physical space; no cosmological claim is made.
PROVING the geometry is coherent (all γ_n real ⟺ spectrum real ⟺ RH) is the OPEN problem. TNFR formalises it as Conjecture T-HP and supplies RH-equivalent coherence/positivity diagnostics (P16 Li–Keiper, P26 Lyapunov-spectral), but the residual oscillation S(T) = (1/π)·arg ζ(½+iT) — visible here as the slow N^{-1/2} convergence — is the genuine obstruction. The TNFR-Riemann program is PAUSED at this boundary; G4 = RH remains open. Coincidence of names ("coherence", "sense") with the TNFR metrics C(t), Si is suggestive but is NOT a proof: see the closing notes.
#!/usr/bin/env python3
"""
Example 95 — Primes from Spectral Waves
========================================
Reconstructs the prime-counting staircase ψ(x) as a superposition of
spectral waves — one per Riemann zero — making tangible the duality
prime positions ⟷ spectral geometry {γ_n}.
Physics
-------
The Riemann–von Mangoldt explicit formula expresses the prime staircase
ψ(x) = Σ_{p^k ≤ x} log p as
ψ(x) = x − Σ_n x^{ρ_n}/ρ_n − log(2π) − ½ log(1 − x^{-2})
with ρ_n = ½ + i·γ_n the non-trivial zeros. Each oscillatory term
x^{ρ_n}/ρ_n = (√x / |ρ_n|) · exp(i·γ_n·log x − i·arg ρ_n)
is a WAVE of angular frequency γ_n in the coordinate log x and amplitude
√x/|ρ_n|. Primes appear where they do because their distribution is the
Fourier synthesis of these spectral waves: the "coherence flow ordering
itself" is literally the superposition of e^{i·γ_n·log x}.
Spectral coherence = Riemann Hypothesis
---------------------------------------
RH states Re(ρ_n) = ½ for EVERY zero. Then every wave has amplitude
√x/|ρ_n| — uniformly bounded — and the staircase stays coherent
(|ψ(x) − x| = O(√x log²x)). A single off-critical-line zero at
Re = ½ + δ would inject a wave of amplitude x^{½+δ} that grows faster
and DESTROYS the bound. So RH ⟺ "all spectral waves equally bounded"
⟺ the spectral geometry is maximally coherent.
Experiments
-----------
1. Staircase synthesis: ψ(x) reconstructed from N = 5, 20, 80 zeros
2. Wave decomposition: each zero as a bounded wave (√x/|ρ_n|, freq γ_n)
3. Coherence ⟺ RH: on-line (Re=½) stays √x-bounded; off-line (Re=0.7)
error grows like x^{0.7} — incoherent
Honest scope
------------
This REPRODUCES the classical explicit formula (Riemann 1859, von
Mangoldt 1895); it is NOT a TNFR discovery. TNFR's contribution is to
realise the prime-side spectrum as a self-adjoint prime-ladder
Hamiltonian (P14, spectrum {k·log p}) and verify the identity to machine
precision (P15). The "spectral geometry" is the geometry of an operator
on the critical line — NOT physical space; no cosmological claim is made.
PROVING the geometry is coherent (all γ_n real ⟺ spectrum real ⟺ RH) is
the OPEN problem. TNFR formalises it as Conjecture T-HP and supplies
RH-equivalent coherence/positivity diagnostics (P16 Li–Keiper, P26
Lyapunov-spectral), but the residual oscillation S(T) = (1/π)·arg ζ(½+iT)
— visible here as the slow N^{-1/2} convergence — is the genuine
obstruction. The TNFR-Riemann program is PAUSED at this boundary; G4 = RH
remains open. Coincidence of names ("coherence", "sense") with the TNFR
metrics C(t), Si is suggestive but is NOT a proof: see the closing notes.
References
----------
- theory/TNFR_NUMBER_THEORY.md §10 (prime path graphs, Riemann link)
- theory/TNFR_RIEMANN_RESEARCH_NOTES.md §13septies (Conjecture T-HP)
- src/tnfr/riemann/analytic_continuation.py (P13 explicit formula)
- src/tnfr/riemann/prime_ladder_hamiltonian.py (P14 spectrum {k log p})
- AGENTS.md §"TNFR-Riemann Program Overview" (status, open G4)
"""
import math
import os
import sys
sys.path.insert(0, os.path.join(os.path.dirname(__file__), "..", "..", "src"))
from tnfr.riemann.analytic_continuation import (
fetch_riemann_zeros,
reconstruct_psi_via_explicit_formula,
)
# ============================================================================
# EXPERIMENT 1: Staircase synthesis from spectral waves
# ============================================================================
def experiment_1_staircase_synthesis():
"""Reconstruct ψ(x) from increasing numbers of spectral modes."""
print("=" * 72)
print("EXPERIMENT 1: Prime Staircase from Spectral Waves")
print("=" * 72)
print()
print("ψ(x) = x − Σ_n x^{ρ_n}/ρ_n − log(2π) − ½log(1−x^{-2})")
print("Each Riemann zero adds one wave. More zeros → sharper staircase.")
print()
xs = [10.5, 12.5, 16.5, 18.5, 22.5, 28.5]
zeros80 = fetch_riemann_zeros(80)
res = {
n: reconstruct_psi_via_explicit_formula(xs, zeros=list(zeros80[:n]))
for n in (5, 20, 80)
}
print(f"{'x':>6} {'ψ_true':>9} {'N=5':>8} {'N=20':>8} {'N=80':>8}")
print("-" * 50)
for i, x in enumerate(xs):
print(
f"{x:>6.1f} {res[80].psi_classical[i]:>9.3f} "
f"{res[5].psi_explicit[i]:>8.3f} "
f"{res[20].psi_explicit[i]:>8.3f} "
f"{res[80].psi_explicit[i]:>8.3f}"
)
print()
for n in (5, 20, 80):
max_err = float(max(res[n].abs_error))
print(f" N={n:>3} zeros → max reconstruction error = {max_err:.3f}")
print()
print("Error shrinks with more spectral modes (slowly, ~N^{-1/2} — the")
print("explicit formula converges only conditionally). Primes ARE the")
print("synthesis of these waves.")
print()
# ============================================================================
# EXPERIMENT 2: Each zero is a bounded spectral wave
# ============================================================================
def experiment_2_wave_decomposition():
"""Show each zero as a wave: amplitude √x/|ρ|, frequency γ_n."""
print("=" * 72)
print("EXPERIMENT 2: Each Zero is a Spectral Wave")
print("=" * 72)
print()
print("x^{ρ}/ρ = (√x/|ρ|)·exp(i·γ·log x − i·arg ρ) with ρ = ½ + i·γ")
print("→ amplitude √x/|ρ|, angular frequency γ in the log-x coordinate.")
print()
x = 100.5
zeros = fetch_riemann_zeros(8)
sqrt_x = math.sqrt(x)
log_x = math.log(x)
print(f"At x = {x}: √x = {sqrt_x:.3f}, log x = {log_x:.4f}")
print()
print(
f"{'n':>3} {'γ_n':>9} {'|ρ_n|':>8} {'amplitude':>10}"
f" {'phase γ·logx':>12}"
)
print("-" * 54)
for n, rho in enumerate(zeros, start=1):
gamma = float(rho.imag)
mod = abs(complex(0.5, gamma))
amplitude = 2.0 * sqrt_x / mod # factor 2 from ρ + conjugate
phase = (gamma * log_x) % (2 * math.pi)
print(
f"{n:>3} {gamma:>9.4f} {mod:>8.4f} {amplitude:>10.4f}"
f" {phase:>12.4f}"
)
print()
print("Every amplitude scales as √x (NOT faster) because Re(ρ) = ½.")
print("Lower zeros (small γ) carry the loudest waves; higher zeros add")
print("fine detail. This is the spectral geometry of the primes.")
print()
# ============================================================================
# EXPERIMENT 3: Spectral coherence ⟺ Riemann Hypothesis
# ============================================================================
def experiment_3_coherence_is_rh():
"""On-line zeros stay √x-bounded; off-line zeros grow like x^{Re}."""
print("=" * 72)
print("EXPERIMENT 3: Spectral Coherence ⟺ Riemann Hypothesis")
print("=" * 72)
print()
print("RH: Re(ρ) = ½ for ALL zeros ⟹ every wave bounded by √x/|ρ| ⟹")
print("coherent staircase. A hypothetical off-line zero (Re = 0.7)")
print("injects a wave of amplitude x^{0.7} that breaks the bound.")
print()
xs = [20.5, 50.5, 100.5, 200.5]
zeros = fetch_riemann_zeros(40)
on = reconstruct_psi_via_explicit_formula(xs, zeros=list(zeros))
off_zeros = [complex(0.7, r.imag) for r in zeros]
off = reconstruct_psi_via_explicit_formula(xs, zeros=off_zeros)
print(
f"{'x':>7} {'ψ_true':>9} {'err ON (Re=½)':>13}"
f" {'err OFF (Re=.7)':>15} {'√x':>7} {'x^0.7':>8}"
)
print("-" * 72)
for i, x in enumerate(xs):
print(
f"{x:>7.1f} {on.psi_classical[i]:>9.2f} "
f"{on.abs_error[i]:>13.3f} {off.abs_error[i]:>15.3f} "
f"{x ** 0.5:>7.2f} {x ** 0.7:>8.2f}"
)
on_ratio = on.abs_error[-1] / on.abs_error[0]
off_ratio = off.abs_error[-1] / off.abs_error[0]
print()
print(
f" Error growth (x: 20→200): ON-line ×{on_ratio:.1f}"
f" vs OFF-line ×{off_ratio:.1f}"
)
print()
print("On-line error tracks √x (coherent). Off-line error tracks x^{0.7}")
print("(incoherent — diverges faster). RH ⟺ the spectral geometry is")
print("maximally coherent: all waves share the same √x envelope.")
print()
def main():
print()
print(" TNFR Example 95: Primes from Spectral Waves")
print(" prime positions ⟷ spectral geometry {γ_n}")
print(" ==========================================")
print()
experiment_1_staircase_synthesis()
experiment_2_wave_decomposition()
experiment_3_coherence_is_rh()
print("=" * 72)
print("HONEST SCOPE & THE COHERENCE QUESTION")
print("=" * 72)
print()
print("This demo REPRODUCES the classical Riemann–von Mangoldt explicit")
print("formula. TNFR realises the prime side as a self-adjoint")
print("prime-ladder Hamiltonian (P14, spectrum {k·log p}) and verifies")
print("the identity to machine precision (P15).")
print()
print("RH is literally a COHERENCE statement: the spectral geometry is")
print("coherent ⟺ every γ_n is real ⟺ the operator is self-adjoint ⟺")
print("all waves share the √x envelope (Experiment 3).")
print()
print("Is this what TNFR's C(t) / Sense Index (Si) can prove? The name")
print("coincidence is suggestive but is NOT itself a proof:")
print(" - C(t) = 1/(1 + mean|ΔNFR| + mean|dEPI|) and Si are metrics of")
print(" a network's dynamic coherence, with specific formulas.")
print(" - RH-coherence is the reality of an operator's spectrum.")
print(" - Equating them is exactly Conjecture T-HP (§13septies): does a")
print(" canonical operator built from the tetrad + (φ,γ,π,e) + U1-U6")
print(" have spectrum {γ_n}? That would make RH a TNFR coherence")
print(" theorem.")
print(" - TNFR already pursued this: P16 (Li–Keiper positivity) and P26")
print(" (Lyapunov-spectral positivity) are RH-equivalent coherence")
print(" diagnostics. The residual S(T) = (1/π)·arg ζ(½+iT) — the slow")
print(" N^{-1/2} convergence seen in Experiment 1 — is the genuine")
print(" obstruction (lives in Fix(S_n)^⊥, unreachable by the current")
print(" canonical constructions).")
print()
print("Verdict: the intuition 'coherence proves the geometry' names the")
print("RIGHT program (T-HP), but the current C(t)/Si metrics do NOT close")
print("it. G4 = RH remains open; the program is paused at this boundary.")
print()
if __name__ == "__main__":
main()