Arithmetic Functions: Multiplicativity and Möbius Function, Assignments of Mathematics

A portion of a university mathematics course, specifically from umass amherst math 471 in the fall of 2006. It covers the topic of arithmetic functions, focusing on multiplicative functions and the möbius function. Definitions, examples, and problems related to arithmetic functions, divisor sums, and the möbius function.

Typology: Assignments

Pre 2010

Uploaded on 08/18/2009

koofers-user-czm
koofers-user-czm 🇺🇸

10 documents

1 / 2

Toggle sidebar

This page cannot be seen from the preview

Don't miss anything!

bg1
UMASS AMHERST MATH 471 FALL 2006, F. HAJIR
HOMEWORK 9: ARITHMETIC FUNCTIONS
Let us make a couple of definitions first. An arithmetic function is a function f:Z+C
where Z+is the set of positive integers. For example, the Euler phi function ϕ:Z+Z+
sending mto ϕ(m) = |(Z/mZ)×|is an arithmetic function. We say that an arithmetic
function fis multiplicative if
f(mn) = f(m)f(n) for all coprime integers m, n Z+.
Please note the condition that the multiplicativity condition f(mn) = f(m)f(n) does not
need to hold for all pairs m, n, only for those that satisfy gcd(m, n) = 1. One way to think of
this is that a multiplicative function fis “free” to hold whatever values it wants on the prime
powers m=pt, but once those are chosen, then all the remaining values are determined by
the fact that each integer is (uniquely) a product of primes.
Given an arithmetic function f:Z+C, we construct the divisor-sum function of f,
called F, by
F(m) = X
d|m
f(d).
Here, by convention, the notation d|m indicates that the sum is over the set of positive
divisors dof m.
Some other standard arithmetic functions are the nu-function ν(m) = |Div+(m)|which
counts the number of divisors of mand the sigma-function σ(m) = Pd|mdwhich gives the
sum of the divisors of m.
A very useful (but at first sight really weird) function is the obius mu (pronounced
“mew”) function µ. It can be defined as follows. We have µ(1) = 1 (see problem 1 below
for why this is a popular choice) and put µ(2) = µ(3) = µ(5) = µ(7) = µ(11) = · · · =1,
in fact let’s put µ(p) = 1 for every prime p. What about prime powers? Here you might
be surprised by the choice I am about to make: If t > 1, and pis a prime, we define
µ(pt) = 0. So, µ“kills” higher prime powers. Finally, we define µon the rest of the integers
by multiplicativity. To put this all together, the definition of µis as follows. First, µ(m) = 1
if m= 1, and µ(m) = (1)rif m=p1· · · pris the product of rdistinct primes p1, . . . , pr.
But if mis divisible by the square of any prime, then µ(m) = 0. You’ll see below why this
is a useful function.
1. Show that if fis a multiplicative arithmetic function, then f(1) = 1.
2. Suppose Div+(m) = {1dm|dis a divisor of m}is the set of positive divisors of
an integer m. Show that if m, n are coprime, then
Div+(mn) = {dd0|dDiv+(m) and d0Div+(n)}.
Hint: Show that the LHS is contained in the RHS and versa vice.
1
pf2

Partial preview of the text

Download Arithmetic Functions: Multiplicativity and Möbius Function and more Assignments Mathematics in PDF only on Docsity!

UMASS AMHERST MATH 471 FALL 2006, F. HAJIR

HOMEWORK 9: ARITHMETIC FUNCTIONS

Let us make a couple of definitions first. An arithmetic function is a function f : Z+^ → C where Z+^ is the set of positive integers. For example, the Euler phi function ϕ : Z+^ → Z+ sending m to ϕ(m) = |(Z/mZ)×| is an arithmetic function. We say that an arithmetic function f is multiplicative if

f (mn) = f (m)f (n) for all coprime integers m, n ∈ Z+.

Please note the condition that the multiplicativity condition f (mn) = f (m)f (n) does not need to hold for all pairs m, n, only for those that satisfy gcd(m, n) = 1. One way to think of this is that a multiplicative function f is “free” to hold whatever values it wants on the prime powers m = pt, but once those are chosen, then all the remaining values are determined by the fact that each integer is (uniquely) a product of primes. Given an arithmetic function f : Z+^ → C, we construct the divisor-sum function of f , called F , by

F (m) =

d|m

f (d).

Here, by convention, the notation “d|m” indicates that the sum is over the set of positive divisors d of m. Some other standard arithmetic functions are the nu-function ν(m) = |Div+(m)| which counts the number of divisors of m and the sigma-function σ(m) =

d|m d^ which gives the sum of the divisors of m. A very useful (but at first sight really weird) function is the M¨obius mu (pronounced “mew”) function μ. It can be defined as follows. We have μ(1) = 1 (see problem 1 below for why this is a popular choice) and put μ(2) = μ(3) = μ(5) = μ(7) = μ(11) = · · · = −1, in fact let’s put μ(p) = −1 for every prime p. What about prime powers? Here you might be surprised by the choice I am about to make: If t > 1, and p is a prime, we define μ(pt) = 0. So, μ “kills” higher prime powers. Finally, we define μ on the rest of the integers by multiplicativity. To put this all together, the definition of μ is as follows. First, μ(m) = 1 if m = 1, and μ(m) = (−1)r^ if m = p 1 · · · pr is the product of r distinct primes p 1 ,... , pr. But if m is divisible by the square of any prime, then μ(m) = 0. You’ll see below why this is a useful function.

  1. Show that if f is a multiplicative arithmetic function, then f (1) = 1.
  2. Suppose Div+(m) = { 1 ≤ d ≤ m | d is a divisor of m} is the set of positive divisors of an integer m. Show that if m, n are coprime, then

Div+(mn) = {dd′^ | d ∈ Div+(m) and d′^ ∈ Div+(n)}. Hint: Show that the LHS is contained in the RHS and versa vice.

1

2 HOMEWORK 9: ARITHMETIC FUNCTIONS

  1. Show that if f is a multiplicative arithmetic function, with divisor-sum function F , then F is also a multiplicative arithmetic function. Hint: It boils down to justifying  

d|m

f (d)

d′|n

f (d′)

d|m

d′|n

f (d)f (d′) =

e|mn

f (e),

for which verification you use 2.

  1. In this problem, we will prove that for an integer m ≥ 1,

(1)

d|m

ϕ(d) = m,

where, as before, the notation “d|m” indicates that the sum is over the set of positive divisors d of m. Let Φ(m) =

d|m ϕ(d) be the divisor-sum function of^ ϕ.^ We want to prove that Φ(m) = m. (i) Verify the formula (1) for 1 ≤ m ≤ 10, recall that ϕ(1) = 1 by definition. (ii) Verify that Φ(m) = m when m = pt^ is a power of a prime p. (Don’t forget the all-purpose identity (x − 1)(1 + x + x^2 + · · · + xt) = xt+1^ − 1). (iii) Prove that Φ(m) = m holds for all m by using (ii) and Problem 3.

  1. (The M¨obius Inversion Formula) Suppose f is a multiplicative arithmetic function, with divisor-sum function F , so that F (m) =

d|m f^ (d).^ Prove that f (m) =

d|m

μ(m/d)F (d)

by following the steps below. Let g(m) =

d|m μ(m/d)F^ (d). (i) Check that g is a multiplicative function. (ii) Check that g(pt) = f (pt) for a prime p and an integer t ≥ 0. (iii) Use (i) and (ii) to conclude that g(m) = f (m) for all m.

  1. Show that for m ≥ 1, ϕ(m) = m

d|m μ(d)/d.

  1. (i) Show that if m = pt 11 · · · pt rr is the prime factorization of m, then ν(m) = (t 1 +
  1. · · · (tr + 1). Hint: you don’t need any fancy machinery for this, just consider the prime factorization of the possible divisors of m and use the “fundamental counting principle”). (ii) Show that ν(m) is odd if and only if m is a perfect square. (Again, you might want to think about this in elementary terms).
  1. Show that for m ≥ 1, (i)

d|m μ(n/d)ν(d) = 1, (ii)

d|m μ(m/d)σ(d) =^ m.