TheoremComplete

Perron Formula

The Perron formula is a cornerstone of analytic number theory, providing an integral representation that recovers summatory functions from Dirichlet series via contour integration.

TheoremPerron's Formula

Let F(s)=βˆ‘n=1∞an/nsF(s) = \sum_{n=1}^{\infty} a_n/n^s be a Dirichlet series converging absolutely for β„œ(s)>Οƒ0\Re(s) > \sigma_0. For x>0x > 0 not an integer, c>Οƒ0c > \sigma_0, and T>0T > 0:

βˆ‘n≀xan=12Ο€i∫cβˆ’iTc+iTF(s)xssds+O(βˆ‘n=1∞∣an∣ncmin⁑(1,xT∣log⁑(x/n)∣))\sum_{n \leq x} a_n = \frac{1}{2\pi i} \int_{c-iT}^{c+iT} F(s) \frac{x^s}{s} ds + O\left(\sum_{n=1}^{\infty} \frac{|a_n|}{n^c} \min\left(1, \frac{x}{T|\log(x/n)|}\right)\right)

As Tβ†’βˆžT \to \infty, the error term vanishes, giving:

βˆ‘n≀xan=12Ο€i∫cβˆ’i∞c+i∞F(s)xssds\sum_{n \leq x} a_n = \frac{1}{2\pi i} \int_{c-i\infty}^{c+i\infty} F(s) \frac{x^s}{s} ds
ExampleRecovering the Prime Counting Function

For F(s)=βˆ’ΞΆβ€²(s)/ΞΆ(s)=βˆ‘n=1βˆžΞ›(n)/nsF(s) = -\zeta'(s)/\zeta(s) = \sum_{n=1}^{\infty} \Lambda(n)/n^s, Perron's formula gives:

βˆ‘n≀xΞ›(n)=12Ο€i∫cβˆ’i∞c+i∞(βˆ’ΞΆβ€²(s)ΞΆ(s))xssds\sum_{n \leq x} \Lambda(n) = \frac{1}{2\pi i} \int_{c-i\infty}^{c+i\infty} \left(-\frac{\zeta'(s)}{\zeta(s)}\right) \frac{x^s}{s} ds

Shifting the contour left and collecting residues yields information about ψ(x)=βˆ‘n≀xΞ›(n)\psi(x) = \sum_{n \leq x} \Lambda(n), the Chebyshev psi function.

DefinitionMellin Transform Connection

The Perron formula is essentially an inverse Mellin transform. For f(x)=βˆ‘n≀xanf(x) = \sum_{n \leq x} a_n:

  • Mellin transform: M[f](s)=∫0∞f(x)xβˆ’sβˆ’1dx\mathcal{M}[f](s) = \int_0^{\infty} f(x) x^{-s-1} dx relates to F(s)/sF(s)/s
  • Inverse: Perron's formula recovers f(x)f(x) from F(s)F(s)

This perspective connects Dirichlet series to transform theory.

ExampleAsymptotic Analysis

The power of Perron's formula lies in contour shifting. The integral:

12Ο€i∫cβˆ’i∞c+i∞΢(s)xssds\frac{1}{2\pi i} \int_{c-i\infty}^{c+i\infty} \zeta(s) \frac{x^s}{s} ds

can be shifted left past s=1s=1 (pole of ΞΆ\zeta), picking up residue xx, giving:

βˆ‘n≀x1=⌊xβŒ‹β‰ˆx+(errorΒ termsΒ fromΒ shiftedΒ integral)\sum_{n \leq x} 1 = \lfloor x \rfloor \approx x + \text{(error terms from shifted integral)}
Remark

The error terms from Perron's formula depend on:

  1. The location of zeros/poles of F(s)F(s) (determines how far we can shift the contour)
  2. Growth estimates of F(s)F(s) on vertical lines (bounds the shifted integral)

For ΞΆ(s)\zeta(s), the Riemann Hypothesis (all non-trivial zeros have β„œ(s)=1/2\Re(s) = 1/2) dramatically improves error bounds.

Perron's formula transforms the problem of understanding sums of arithmetic functions into a problem of analyzing analytic properties (zeros, poles, growth) of Dirichlet seriesβ€”the essence of analytic number theory.