« first day (1445 days earlier)      last day (3579 days later) » 
03:00 - 22:0022:00 - 00:00

10:00 PM
I hate it if it just comes from algebraic manipulations. Combinatorial recurrences are fine by me.
 
@BalarkaSen Have you decided what unis you will be applying to?
 
@JasperLoy nah. not right now!
 
$$t(\text{n$\_$},1)=1;$$
$$t(\text{n$\_$},\text{k$\_$})\text{:=}t(n,k)=\text{If}[n\geq k,t(n-1,k-1)+t(n-1,k),0];$$

$$t(\text{n$\_$},1)=1;$$
$$t(\text{n$\_$},\text{k$\_$})\text{:=}t(n,k)=\text{If}\left[n\geq k,\sum _{i=1}^{n-1} t(n-i,k-1),0\right];$$

$$\begin{array}{llllllllll}
1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\
1 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\
1 & 2 & 1 & 0 & 0 & 0 & 0 & 0 & 0 & 0 \\
1 & 3 & 3 & 1 & 0 & 0 & 0 & 0 & 0 & 0 \\
1 & 4 & 6 & 4 & 1 & 0 & 0 & 0 & 0 & 0 \\
1 & 5 & 10 & 10 & 5 & 1 & 0 & 0 & 0 & 0 \\
 
University education should not be so expensive. It is very sad.
 
@MatsGranvik The first recurrence is just what I posted above.
And the second comes merely from adding stuffs.
What's the big deal?
 
10:04 PM
@BalarkaSen Yes it is. But this is key.
 
@robjohn Do US universities usually have a Calculus 1,2,3 sequence?
 
@MatsGranvik OK. So you have analogies or something?
 
@BalarkaSen what analogy?
 
@MatsGranvik For divisor functions, like you said...?
 
Never mind. I misunderstood.
 
10:05 PM
What would happen if all SE chat rooms were combined into one?
 
$$t(\text{n$\_$},1)=1;$$

$$t(\text{n$\_$},\text{k$\_$})\text{:=}t(n,k)=\text{If}\left[n\geq k,\sum _{i=1}^{k-1} (t(n-i,k-1)-t(n-i,k)),0\right];$$

http://i.stack.imgur.com/sggys.png
@BalarkaSen
 
And is there any closed form for $t([], [])$?
As defined above?
 
For the divisor function? No not that I know of but I can mix in complex numbers for summatory variants and get asymptotics.
 
@MatsGranvik Oh so your $t$ is the divisor function $\tau$?
 
@BalarkaSen No not so, the $t$ is just the values in the table. Tau is the row sums.
 
10:10 PM
@MatsGranvik Row sums?
 
https://oeis.org/A000005
1, 2, 2, 3, 2, 4, 2, 4, 3, 4, 2, 6, 2, 4, 4, 5, 2, 6, 2, 6, 4, 4, 2, 8, 3, 4, 4, 6, 2, 8, 2, 6, 4, 4, 4, 9, 2, 4,

is the row sums of the plot which is the table https://oeis.org/A051731
@BalarkaSen
 
So that's what you mean by row sum.
 
Yes, the recurrence merely describes the plot, not the row sums.
 
So you are claiming that $\tau(n) = \sum_{k = 1}^n t(n, k)$?
 
10:15 PM
@MatsGranvik And do the values of t(n, k) and A051731 match?
 
@BalarkaSen precisely, they match yes.
 
@MatsGranvik so the claim is that t(n, k) defined as above satisfied t(n, k) = 1 if k|n and 0 otherwise.
 
@BalarkaSen You can put it that way also.
 
do you have a proof?
 
Jeffrey O Shallit gave a proof that I have somewhere in my mail inbox.
 
10:18 PM
@MatsGranvik aha.
2
 
@MatsGranvik Who is he?
 
I know Jeffrey. He is a regular contributor in OEIS
 
What is the use of OEIS?
 
I think he is also the editor of a journal called integers.
 
@JasperLoy Give a sequence get a bunch of properties.
@MatsGranvik Yes, indeed.
 
10:19 PM
@BalarkaSen I see. I don't think I will ever need to use it.
 
@JasperLoy you won't. not if you are interested in nt, decimal expansions of numbers, and cheating some puzzle problems.
and many other things i am missing.
 
Jeffrey O'Shallit's proof:

Here's the proof.

Now (sum from i=1 to k-1 of T(n-i,k-1)) is just the number of elements in
{n-k+1,...,n-1) that are divisible by k-1. This is always equal to 1,
because in k-1 consecutive
integers, exactly one will be divisible by k-1.

On the other hand, (sum from i = 1 to k-1 of T(n-i,k)) is equal to
either 0 or 1, because in
k-1 consecutive integers, either 0 or 1 of them are divisible by k.
We get 1 except if
all are not divisible by k, which means the next integer /is/ divisible
 
OK. That wasn't much of a big deal. So you have a claim and you have a proof.
My question : Are there any interesting consequences?
 
@BalarkaSen Do you know the Möbius function?
 
Yes.
I am familiar with it.
In fact I am familiar with it very well. Gave me a hard time when doing some asymptotics for the PNT.
=P
 
10:24 PM
Ok, by shifting the second columns recurrence index one step down, you can calculate the Möbius function as a matrix inverse from an arbitrary number sequence. But of course I told you this yesterday. @BalarkaSen
 
what do you mean by taking the index one step down?
 
$$\begin{align}
T(n,1) &= (n-1)\text{th digit of }\pi, \\
T(n,2) &= T(n,1) - T(n-1,2), \\
\text{for } k>2, T(n,k) &= \sum\limits_{i=1}^{k-1} T(n-i,k-1)-\sum\limits_{i=1}^{k-1} T(n-i,k)
\end{align}$$
 
Ah. OK.
 
@JasperLoy yes
 
So you are defining $T(n, k)$ as that for a given arbitrary number sequence $T(n, 1)$. @Mats
 
10:27 PM
@blue I am thinking of getting Marsden's Calculus 1,2,3.
 
@BalarkaSen The matrix inverse of that table $T(n,k)$
is the möbius function.
 
That's intriguing.
And I presume this is a conjecture of yours?
 
@blue I am very sad that I am mad.
 
Yes, it is but Thomas Andrews has almost proven it. But I don't understand it. @BalarkaSen
 
@MatsGranvik Where is his proof? I might get a peek at it.
 
10:30 PM
Here:
http://math.stackexchange.com/a/268213/8530
 
@anorton Long time no see
 
@JasperLoy Same to you! How are you?
 
@anorton Still bad. I am very sad that I am mad.
 
:( Sorry to hear that.
 
Interesting idea. I'd have to write it up later explicitly to understand what's he trying to say but I think the key is that $\sum_{d|n} \mu(d) = 0$ for $n > 1$ and $1$ otherwise. You'll have to start thinking about $A(1, 0, 0, \cdots )$. I can't imagine what shape it takes.
 
10:36 PM
Hi, here is a somewhat interesting problem on an infinite product of matrices: mathb.in/18525?key=ad31a01ab5b80a9685525eeefd3760aff976556a , I have tried to use matrix norms, to look into spectral radius.. What else could I try?
 
10:49 PM
@JasperLoy ?
 
@blue Just felt like saying.
 
Hmm.
@blue If L/K is a galois extension then it turns out that O_L is a Gal(L/K)-module over O_K.
 
yes, that's part of galois module theory
 
galois module theory?
 
there is much work on figuring out when O_L is free over the associated order A_L/K
good survey by lara thomas iirc
 
10:53 PM
@blue what the hell is A_L?
 
${\cal A}_{L/K}=\{x\in K[G_{L/K}]:x{\cal O}_L\subseteq{\cal O}_L\}$
 
The fractional ideal, aka.
 
no, it's an algebra
 
oh right.
 
it can happen to strictly contain the group ring ${\cal O}_K[G_{L/K}]$
 
10:54 PM
i was deceived by the inclusion.
@blue and what's really the motivation for all these stuffs?
 
apparently the stuff is in and of itself motivation
anyway I have to go, library closing
 
byes.
hmm. the ramification groups defined in Thomas stirs something up in the memory.
of course. they are the automorphisms defined in the correspondence between cohomology groups. is that just a coincidence?
 
11:42 PM
back
dunno any cohomology
 
hi @LoganM what's new in the world of Physics?
 
@skullpatrol I am sad that I am mad.
 
I'm mad too.
 
Really?
 
yep
 
11:54 PM
They say misery loves company.
 
It does
 
03:00 - 22:0022:00 - 00:00

« first day (1445 days earlier)      last day (3579 days later) »