Author 
Message 
eugene science forum Guru
Joined: 24 Nov 2005
Posts: 331

Posted: Mon Jul 10, 2006 9:35 am Post subject:
Dense subset of C



It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
Thanks 

Back to top 


William Elliot science forum Guru
Joined: 24 Mar 2005
Posts: 1906

Posted: Mon Jul 10, 2006 9:51 am Post subject:
Re: Dense subset of C



On Mon, 10 Jul 2006, eugene wrote:
Quote:  It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
How much thought have you given this? 
Did you look for any counter example? 

Back to top 


eugene science forum Guru
Joined: 24 Nov 2005
Posts: 331

Posted: Mon Jul 10, 2006 10:01 am Post subject:
Re: Dense subset of C



William Elliot wrote:
Quote:  On Mon, 10 Jul 2006, eugene wrote:
It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
How much thought have you given this?
Did you look for any counter example?

Yes, you are right? Sorry, i really didn't thought much on this.
We may put v = 3 * i and w = 3 for example and this set doesn't even
have accumulation points.
In fact, this question was arised from the problem:
f: C > C analytic and v, w in C such that v/w is not real such that
f(z + v) = f(z) = f(z + w) for all z in C. Prove that f is contsant.
There should probably be another apprach for this.
Thanks and sorry 

Back to top 


Lasse science forum addict
Joined: 29 Apr 2005
Posts: 71

Posted: Mon Jul 10, 2006 10:04 am Post subject:
Re: Dense subset of C



eugene wrote:
Quote:  It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?

HINT: Think of two such complex numbers, and try to work out what the
given set will look like in the complex number plane. 

Back to top 


Lasse science forum addict
Joined: 29 Apr 2005
Posts: 71

Posted: Mon Jul 10, 2006 10:08 am Post subject:
Re: Dense subset of C



Quote:  In fact, this question was arised from the problem:
f: C > C analytic and v, w in C such that v/w is not real such that
f(z + v) = f(z) = f(z + w) for all z in C. Prove that f is contsant.
There should probably be another apprach for this.

Hint: any entire function which is bounded is also constant. 

Back to top 


Lee Rudolph science forum Guru
Joined: 28 Apr 2005
Posts: 566

Posted: Mon Jul 10, 2006 2:33 pm Post subject:
Re: Dense subset of C



"eugene" <jane1806@rambler.ru> writes:
Quote:  It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?

Did you actually *think* about this question for, say, *one second*?
Maybe try, y'know, a few *examples* of v and w? Say, w = 1, and v...
well, gosh, what's some nonreal complex number? i wonder.
Lee Rudolph 

Back to top 


eugene science forum Guru
Joined: 24 Nov 2005
Posts: 331

Posted: Mon Jul 10, 2006 2:37 pm Post subject:
Re: Dense subset of C



Lee Rudolph wrote:
Quote:  "eugene" <jane1806@rambler.ru> writes:
It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
Did you actually *think* about this question for, say, *one second*?
Maybe try, y'know, a few *examples* of v and w? Say, w = 1, and v...
well, gosh, what's some nonreal complex number? i wonder.
Lee Rudolph

I gave example some posts above. 

Back to top 


Lee Rudolph science forum Guru
Joined: 28 Apr 2005
Posts: 566

Posted: Mon Jul 10, 2006 2:48 pm Post subject:
Re: Dense subset of C



"eugene" <jane1806@rambler.ru> writes:
Quote:  Lee Rudolph wrote:
"eugene" <jane1806@rambler.ru> writes:
It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
Did you actually *think* about this question for, say, *one second*?
Maybe try, y'know, a few *examples* of v and w? Say, w = 1, and v...
well, gosh, what's some nonreal complex number? i wonder.
Lee Rudolph
I gave example some posts above.

So you did; apparently the post I replied to arrived here before
the various replies to it.
Anyway, to expand on Lasse's second hint: first, with v and w as
above, L = { m v + n w  m,n \in Z} is NEVER dense in C. It is, in
fact, a discrete subgroup of the abelian topological group C (isomorphic
to Z/2Z), and the quotient group C/L is therefore itself both a topological
group (just because L is normal in C) and a 2manifold (because L is
discrete). Consequently C/L is, topologically, a torus. If (as you
asked in another of the latearriving posts) f is a holomorphic
function on C such that f(z+v) = f(z) = f(z+w) for all z in C,
then f descends to a welldefined function g on C/L, which is clearly
continuous since f is. (In fact, there is a natural way to put the
structure of a complex curve on C/L so that g is holomorphic, but we
don't need this.) Because C/L, being homeomorphic to a torus, is compact,
g has a local maximum on C/L. Therefore f has local maxima on C.
Therefore f is constant.
Lee Rudolph 

Back to top 


The World Wide Wade science forum Guru
Joined: 24 Mar 2005
Posts: 790

Posted: Tue Jul 11, 2006 2:45 am Post subject:
Re: Dense subset of C



In article <e8tpbv$mgn$1@panix2.panix.com>,
lrudolph@panix.com (Lee Rudolph) wrote:
Quote:  "eugene" <jane1806@rambler.ru> writes:
Lee Rudolph wrote:
"eugene" <jane1806@rambler.ru> writes:
It is well known fact that if a,b are real and a/b isn't in Q, then the
set { ma + n b  m,n \in Z} is
dense in R(usually it is called Kronecker's theorem  it's weaker
version). I have the following question: if v and w are complex
nonzero numbers such that v/w is nonreal. Is it true that
{ m v + n w  m,n \in Z} is dense in C ?
Did you actually *think* about this question for, say, *one second*?
Maybe try, y'know, a few *examples* of v and w? Say, w = 1, and v...
well, gosh, what's some nonreal complex number? i wonder.
Lee Rudolph
I gave example some posts above.
So you did; apparently the post I replied to arrived here before
the various replies to it.
Anyway, to expand on Lasse's second hint: first, with v and w as
above, L = { m v + n w  m,n \in Z} is NEVER dense in C. It is, in
fact, a discrete subgroup of the abelian topological group C (isomorphic
to Z/2Z), and the quotient group C/L is therefore itself both a topological
group (just because L is normal in C) and a 2manifold (because L is
discrete). Consequently C/L is, topologically, a torus. If (as you
asked in another of the latearriving posts) f is a holomorphic
function on C such that f(z+v) = f(z) = f(z+w) for all z in C,
then f descends to a welldefined function g on C/L, which is clearly
continuous since f is. (In fact, there is a natural way to put the
structure of a complex curve on C/L so that g is holomorphic, but we
don't need this.) Because C/L, being homeomorphic to a torus, is compact,
g has a local maximum on C/L. Therefore f has local maxima on C.
Therefore f is constant.

For those who wish to stay in C, notice C = U_{m,n in Z} (mv +
nw) + K, where K = {av + bw : a, b in [0,1]}. So f(C) = f(K), and
we're done by Liouville. 

Back to top 


Lee Rudolph science forum Guru
Joined: 28 Apr 2005
Posts: 566

Posted: Tue Jul 11, 2006 10:39 am Post subject:
Re: Dense subset of C



The World Wide Wade <waderameyxiii@comcast.remove13.net> writes:
Quote:  In article <e8tpbv$mgn$1@panix2.panix.com>,
lrudolph@panix.com (Lee Rudolph) wrote:
....
Because C/L, being homeomorphic to a torus, is compact,
g has a local maximum on C/L. Therefore f has local maxima on C.
Therefore f is constant.
For those who wish to stay in C, notice C = U_{m,n in Z} (mv +
nw) + K, where K = {av + bw : a, b in [0,1]}. So f(C) = f(K), and
we're done by Liouville.

And for those who wish both to stay in C and to restrict their tools to
the maximum modulus principle (instead of appealing to Liouville), notice
that f has a maximum somewhere on the compact set K (a parallelogram
and its interior) and therefore has a local maximum somewhere in the
interior of the compact set U_{m,n in Z, m\le 1, n\le 1} (mv+nw)+K.
Lee Rudolph 

Back to top 


Google


Back to top 



The time now is Tue Nov 21, 2017 6:22 am  All times are GMT

