Re: the number of divisors, number theory
- From: Narcoleptic Insomniac <i_have_narcoleptic_insomnia@xxxxxxxxx>
- Date: Tue, 05 Dec 2006 16:27:59 EST
On Dec 5, 2006 2:26 PM CT, W. Schols wrote:
is it true and when yes, can someone prove that
2 exp( omega (n)) <= tau (n) < 2 sqrt (n),
with the following definitions,
n is a positive integer
omega (n) denotes the number of distinct prime
factors of n
tau (n) is the number of divisors of n
sqrt (n) is the square root of n.
Can someone help me with the second part of the
statement ? (the first part is clear for me, but I
have problems to prove the second part, that tau(n)
is always smaller than 2 time the sqrt of n).
thanks and greetings, Wim
I'm not sure if this is the proper way to proceed, but
this is what I thought of. I'm going to use elementary
calculus, but maybe induction would work too.
By the fundamental theorem of arithmetic we have
n = (p_1)^(k_1) * ... * (p_r)^(k_r),
for some primes p_i and integers k_i. Recall that
tau(n) = (k_1 + 1) * ... * (k_r + 1),
and that tau is a multiplicative function. Since the
square root is a completely multiplicative function it
suffices to show the inequality
tau(p^k) < 2 sqrt(p^k),
for some prime p and integer k. In other words, we must
show that
k + 1 < 2 * p^(k/2),
for all k >= 1 for any given p. Now define
f(k) := 2 p^(k/2) - k - 1,
then the derivative w.r.t. k is
f'(k) = ln(p) p^(k/2) - 1.
Note that since p >= 2 and fixed, f'(k) is increasing on
the inverval [1,oo). Moreover, f(1) = 2 p^(1/2) - 2 > 0
for all primes p >= 2.
Thus, f(k) > 0 for all p >= 2 on the interval [1, oo), in
other words
2 p^(k/2) - k - 1 > 0
2 p^(k/2) > k + 1
2 sqrt(p^k) > tau(p^k)
for all p >= 2, k >= 1. Unless of course I made a
mistake somewhere up there.
Regards,
Kyle Czarnecki
.
- Follow-Ups:
- Re: the number of divisors, number theory
- From: wschols
- Re: the number of divisors, number theory
- References:
- the number of divisors, number theory
- From: Wim Schols
- the number of divisors, number theory
- Prev by Date: Re: Variation on binomial distribution
- Next by Date: Re: Cantor Confusion
- Previous by thread: the number of divisors, number theory
- Next by thread: Re: the number of divisors, number theory
- Index(es):
Relevant Pages
|