« first day (306 days earlier)      last day (3766 days later) » 

19:16
@alizter hi
@BalarkaSen So i have been interested in proofs of the PNT. What is your most preferred one?
the all-CA proof.
the proofs with real analysis are pretty boring
@BalarkaSen is that the first one?
@Alizter no, i think
@BalarkaSen I am very much intrigued by the sieving properties you can extract from the riemann zeta function.
However it seems to me like a special case
because the sieve is based of factors.
I am trying to understand if Dirchlet series are relevent for prime like sieve generated sequences.
19:19
@Alizter there is no sieve property that can be inferred from zeta
@BalarkaSen When you turn the zeta into a euler product
OK, then?
you are using properties of sieve of Eratosthenes
god that is hard to spell
where?
When you multiply zeta by 1/2^s for example
(1-1/2^s) zeta(s) = something
and you keep going
that is using sieving properties no?
19:22
that's no sieve.
It is. By subtracting you are removing from the list
trivial stuff
sieve theory is something much deep
note that I never said sieve theory
now onto sieve theory
trying to apply methods similar methods to a sieve. What is the best way to prove existence of dirchlet series?
@Alizter what is a "dirichlet series"?
you mean $L(1, \chi)$?
yes cannot spell
19:26
I didn't know that could be done by sieves.
It's just some analytic number theoretic estimations.
hmm. a simple way to put it is: I am looking for a zeta analgoue for a prime sequence analgoue
i wondered what properties must my sequence have
i don't get you.
@BalarkaSen What don't you get?
19:29
"zeta analogue for a prime sequence analogue"
Let's define our "prime sequence analogue"
$\{a_n\}\subset \Bbb N$ and satisfies some properties that I am trying to work out
A sequence that satisfies these unkown properties is the primes
one property could be stated as: There should be a sieve for generation of all $a_n$ below some integer $k$. For the primes this is the sieve of Eratosthenes and we have $\pi(k)$
Now I am guessing educatedly here. A PNT style asymptotic relation for the counting function of $a_n$.
I think there was a result somewhere which stated roughly that most $a_n \sim A k \log k$
@Alizter But what is $a_n$?
another property would be $a_n$ are the coefficients in a euler product which produce the corresponding dirichlet series
and for the priems we have $\zeta$
@BalarkaSen That is what I am trying to work out
how can I calssify such an $a_n$
what are the conditions?
@BalarkaSen some of the ones stated above
19:36
Can't make sense out of any of them, sorry.
I think you are being too vague
@BalarkaSen I will give you an example of such a $a_n$ other than the primes
OK. What is le example?
For primes we say. Start at one. Ignore one. Goto 2. Delete every 2nd element. goto 3 delete every third element on old list. etc.
right?
Consider the same sieve however instead of looking at every third element on the old list do it one the new list.
Right.
@Alizter ?
You've lost me.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
@BalarkaSen sieve
for the primes we can start at one
our sieve ignores 1
19:40
ok it's not working blah
however when we look at 3 we cancel every 3rd element
right?
@Alizter I am familiar with Eratosthenes/
What is your sieve supposed to be?
but this is 3rd in relation to the orignal set right?
what if we did 3rd in relation to the left over elements
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 3 5 7 9 11 13 15
now delete 3rd
1 3 7 8 13 15
now 7
19:41
Yes, this is well known.
@BalarkaSen Ok so are you following now?
Let me recall the name.
lucky numbers I think
I read about it a long time ago.
19:42
yes, @Alizter
lucky numbers
Ok. It is proven (somewhere?) that $a_n \sim n \log n$
which is the analgoue of PNT for lucky numbers
$a_n$ being the lucky numbers?
not sure
and actually alot of theorems about primes are well defined for lucky numbers and apply
19:44
that might be some interesting elementary number theory problem to think about, @Alizter
@BalarkaSen the problem I am trying to get at isnt
what is your problem then?
consider such sequences $a_n$ that also share analoguous properties to the primes
you need to define those, @Alizter
otherwise your question is void
i dunno it might be interesting, but think about it. i encourage you to do so.
@BalarkaSen What properties are needed to make it analogous enough that goldbach conjecture for one sequence is equivalent to goldbach for another
replace goldbach with any theorem of primes
twin prime etc.
19:48
no idea
but do you kind of get what I am trying to get at?
yes, I do.
phew
how did the $c_n$ problem go?
And I'd like to see a proper mathematical formulation plus a solution to your problem, @Alizter
@Alizter I forgot about that stuff :P
I am studying topology right now
Hmmm. Still eagerly awaiting for the post on MHB :P
19:50
maybe after the incoming examinations.
How are they going?
they are on 7th next month
btw, if you have anything else to ask, askaway cause i have some works to do.
@BalarkaSen Nothing much. It was mostly that I am thinking about these days.
OK. And for the proof of PNT study a first few pages of Iwaniec-Kowlaski and the PNT chapter of Titchmarsh
follow my MHB post if nessesary.

« first day (306 days earlier)      last day (3766 days later) »