« first day (2790 days earlier)      last day (2223 days later) » 

4:00 AM
now you're pushing it
i'm this close to calling you a physicist. im good with name calling.
 
@JoeShmo you can also read it to construct the inverse if you're interested
 
are you not aware that you're being rude @Leaky?
 
@LeakyNun This looks good
 
@EricSilva the guy below you isn't
 
but i also need to show D exp has full rank
 
4:01 AM
@JoeShmo He frequents $\hbar$
 
who?
 
@JoeShmo h bar

 The h Bar

General chat for Physics SE (physics.stackexchange.com). For M...
 
Shit dude, if you think that garbage is physics, you don't know anything about analysis nor physics :P
 
ive been doing this for more than 12 hours straight now guys. if there's a joke hiding in there just give it to me straight
 
@JoeShmo What are you trying to do?
 
4:03 AM
homework :D
 
Ocelot actually is a physicist is the point
 
well what is the issue
 
@0celo7 I don't care about analysis
 
@AkivaWeinberger I have more PDE books than physics books
which is to say, I have a lot of both
 
nor physics
 
4:03 AM
Right so it's abundantly clear that none of us actually know how to do this homework problem but we've found a route that'll work probably but might not be the intended solution @JoeShmo
 
@LeakyNun nice, tell me more
 
@JoeShmo Also maybe sleep
 
@JoeShmo i gave you a link to the inverse function if it's of any use
 
nobody knows that intended solution
 
@JoeShmo Look, I can do this. Tell me what it is.
 
4:04 AM
@0celo7 show dim(U(n)) = n^2
real dimension
 
well.... i wanna do it by myself
 
Does the same trick for SO(n) not work?
 
which kinda defeats the purpose of asking here
 
What's the exponential map of $\begin{bmatrix}0&0\\1&0\end{bmatrix}$
 
@0celo7 he doesn't know Lie theory
 
4:04 AM
Legit question
 
but the question asks
Show that the set U(n) of unitary n × n-matrices (those n × n-matrices A with
entries in C such that A¯T A = I, where A¯ is the matrix with entries the complex conjugates
of those of A) is a manifold. What is its dimension?
 
Or triangular matrices in general
 
@AkivaWeinberger it's nil-potent, series terminates, compute by hand
 
The trick for SO(n) is to differentiate the transpose map.
 
@AkivaWeinberger extract the diagaonal part
 
4:05 AM
yeah, im trying to apply the regular value theorem
 
@LeakyNun OK but just tell me
 
its the key to all of them
 
@AkivaWeinberger just WA that
 
@LeakyNun Triangular without a diagonal
@LeakyNun は
 
but to apply the regular value theorem i gotta find the proper codomain,
such that the total derivative has full rank
 
4:06 AM
@AkivaWeinberger it's nilpotent; series terminates; it's a finitary problem; it can be solved using finitely many computations
 
@LeakyNun Right OK but I don't want to
 
but that looks wrong
 
@LeakyNun Really? I wouldn't expect that thing in the top-right
 
nvm, I know where the problem is
 
@0celo7 where is the following?
/giphy trolling
 
4:08 AM
@JoeShmo What do you get when you differentiate the $A\mapsto A^*A$ map?
 
we really need to get proper /giphy integration here guys
 
@AkivaWeinberger $\begin{pmatrix}1&0\\1&1\end{pmatrix}$
 
I was asking him
 
why would you answer the question that was clearly not intended for you
 
...thinking...
 
4:09 AM
 
i saw the answer leaky sneaky
doesn't mean anything though so its all good
nope
 
@0celo7 the physicist to the rescue
@JoeShmo have you done any matrix derivative at all?
 
@JoeShmo Do you have Guillemin and Pollack's book?
Check out page 22
If you apply this reasoning to U(n) you can very likely solve it
 
Does $\exp(A)=I+A$ work for all lower triangular matrices?
I guess I need $A^2=0$
Are they nilpotent?
 
@AkivaWeinberger yes
 
4:11 AM
i dont have his book unfortunately
 
You have the internet
 
but the nilpotent degree depends on how close you are to the diagonal
 
Wait actually I don't have $A^2=0$ always
 
amazon doesn't give page 22
 
just below the diagonal is A^(n-1)=0
 
4:12 AM
Russian servers
 
What if it's zero everywhere except one single cell @LeakyNun
off the diagonal
 
still depends on where it is
 
I think its square is zero then
 
many thanks
 
4:12 AM
hmm, maybe
 
@LeakyNun No 'cause it would multiply with other stuff on its diagonal which gives you zero
 
right
 
is there a name for a matrix $A^2=-I$?
 
@Secret "a matrix whose square is the negative identity"
(useful fact: those matrices are diagonalizable)
 
because I will imagine it will be used when the vector spaces of real matrices is complexificated?
 
4:13 AM
> complexificated
 
Right so now we have a basis for the skew-Hermitian matrices and we know the exponential of them all @LeakyNun
 
to be fair, $\Bbb C$ is a real vector-space
 
so now we can compute the map's derivative?
 
jesus this book is so much better than the excuse-for-notes that we get
 
(This is what I have in mind in that question btw):
In mathematics, the complexification of a vector space V over the field of real numbers (a "real vector space") yields a vector space VC over the complex number field, obtained by formally extending the scaling of vectors by real numbers to include their scaling ("multiplication") by complex numbers. Any basis for V (a space over the real numbers) may also serve as a basis for VC over the complex numbers. == Formal definition == Let V be a real vector space. The complexification of V is defined by taking the tensor product of V with the complex numbers (thought of as a two-dimensional vector space...
 
4:15 AM
I can't believe Ted didn't tell you to get it
 
G&P is reeeeeaaaal good @JoeShmo
 
which book?
 
@LeakyNun Federer
 
@JoeShmo What book
 
@0celo7 federerererererer
 
4:15 AM
he did mention it in passing, but then sent me to his lectures on the subject, which are good. But always prefer a book
 
The thing Leaky linked you?
 
its a classic. i see it come up everywhere
 
Zee
Nobody reads that book...
 
Yeah I see it come up everywhere also even though I've never read it
I probably should at some point
 
4:16 AM
::eye roll::
 
And by "everywhere" I mean "on this chat, occasionally"
 
(not you akiva)
 
Zee
The Federer I mean
 
(aight)
 
::eye roll::
 
4:17 AM
Oh, I've heard bad things about Federer
 
@Zee idk dude im reading it
 
It's a shit book, but you still read it
 
tbf people dont read it cuz it's good
 
@EricSilva In our seminar we need some of the Chapter 5 stuff that's not in Simon. Scary shit.
 
Zee
4:18 AM
people don't read it couse it’s not meant for learning but as a reference
 
That's like saying people don't learn from/read papers.
 
Zee
I don’t know how you made that conclusion but it’s actually kinda true as well
 
dude lots of people learn from papers
 
I think this guy is a sophisticated troll.
 
Zee
Of course people learn from papers otherwise nobody would write papers
 
4:20 AM
I wonder what kind of matrix should I use for "i" for the following:
 
Zee
But those people that learn from those papers, are called EXPERTS
 
dude you dont have to be an expert to learn from a paper
 
Let $M_n(\Bbb{R})$ be the vector space of real matrix. I want to produce the following complexification:
 
how the hell would you become an expert if you didnt read papers
 
Zee
By having an advisor who spoon feeds you papers
 
4:21 AM
ok so @0celo7 what was the question again?
 
$$M = M_1 \otimes I + M_2 \otimes A$$
 
@0celo7 i think youre right
 
@JoeShmo Um. I don't think you can answer my question...
 
but $A^2=-I$ seemed to have more than two solutions
so it is not clear which one I should use
 
Zee
Anyway , I don’t mean to appeal to authority , but one of the best geometric measure theorist alive told me (he was teaching a complex analysis class) “I don’t know anybody who used Federer as a beginner “
 
4:23 AM
Cool story bro
 
Zee
just Trying to save you time...
Also , I got a c in his class, Incase anybody thinks am cool ...
 
Don't worry, nobody thinks that
 
@Secret $M_1+M_2A\mapsto M_1-M_2A$ would be an automorphism I think, so maybe one way to start would be to look for automorphisms of the group of matrices
Involution automorphisms
 
@0celo7 went straight savage mode
 
@0celo7 harsh man
 
Zee
4:25 AM
@0celo7 if nobody thinks am cool , then that kinda makes me cool right ?
Being the uncoolest man alive
 
You're not the uncoolest
 
Akiva: hmm I see, should try that
 
@Secret Or actually the $\otimes$ there makes it confusing
I'm not sure I know what it means
Although:
4
Q: Each automorphism of the matrix algebra is inner.

awllowerExercise I.II.IV in the book Local Representation Theory by J.L.Alperin: Demonstrate that any automorphism of the algebra $M_n(k)$ is inner by using the fact that $M_n(k)$ has a unique simple module. I want only hints, as this seems elementary. Thanks for your attention.

 
Zee
Damn, I wanna be uncooler
 
chat is expelling toxic fumes tonight tbh
 
Zee
4:27 AM
Sorry, I had some beans this morning
Can’t tell if that was so funny or so unfunny ...
 
4:45 AM
There are too many theorems called the "Riesz Representation Theorem"
TOO MANY
 
@XanderHenderson how many
 
$\Huge\text{TOO MANY}$
also, possibly, $\Huge\mathfrak{TOO\ MANY}$
 
TWO MANY
 
Could I say the recurrence relation is basically the same as mathematical induction? I just realize that they're very similar...
 
You could, but I am not sure what you would mean by that...
 
4:53 AM
I know at most 3
I feel bad
how many are there
 
three it too many, but I think that i have encountered four or five in my life
though a few of them are really the same thing
there's the Hilbert space version, and the positive measure version
 
those are the big ones
 
@Niing They definitely are closely related
 
more generally, $L^p \simeq (L^q)^\ast$ sometimes gets that moniker
 
The fact that recursive definitions define functions everywhere is the same as induction, essentially, yeah
 
4:55 AM
yes, those are the 3 I know
 
Well maybe strong induction, depending
though strong induction is equivalent to regular old induction
 
Wait I knew the L^p theorem as Riesz theorem without the word representation
 
@AkivaWeinberger that's the tensor product, you need something like that in complexification
presumably that will mean the automorphism will then be applied to the vector space $M_n(\Bbb{R}) \otimes V$
where $V$ is a vector space that includes $A,I, -I$
 
@0celo7 whats the dimension of the hermitian matrices?
 
5:06 AM
1
 
real or complex dimension?
 
either
its a difference of a factor of 2
 
well, the diagonal elements have to be real numbers.
 
why?
over the complex numbers then
 
how are you going to have a hermitian matrix if it's got complex numbers on the diagonal?
 
5:09 AM
(I felt trying to have $V=M_n(\Bbb{C})$ will be a bit excessive, but I need to think about what subspace of $M_n(\Bbb{C})$ I need in order to get $A$)
 
right
so whats the dimension of the complex hermitian matrices?
n^2?
 
(n^2-n)/2 + n = n^2/2
 
(Alternately, I think $M=M_n(\Bbb{R})\oplus AM_n(\Bbb{R})$ might work )
 
yep
so yeah, it's n^2
 
5:11 AM
^over the reals. Over C, n^2
now, how do the unitary matrices relate to hermitian matrices?
 
unitary means $U^\dagger U=1$
You don't need U hermitian, though
 
yes, but im trying to show U_n(C) is a manifold, taking advantage of the regular value theorem
so I want to have phi: U_n(C) -> hermitians
 
Zee
@Daminark something came over me
 
so i need to relate the two spaces somehow
 
Can someone explain this interesting MO answer: mathoverflow.net/a/25231/13542
the claim is that $k[[x,y,z]]$ can inject into $k[[u,v]]$ via :"Take a map from $k[[x,y,z]]$ to $k[[u,v]]$ that sends $x$ to $u$, $y$ to $uv$ and $z$ to $uf(v)$ for some $f\in k[[v]]$ and think about what the kernel is. It isn't hard to see that only for countably many choices of ff can it possibly be zero."
isn't the kernel never trivial? since x=0 makes the image zero
 
5:31 AM
"these results were obtained by the writer before the war"
 
Akiva: (Not guarantee to work, because I don't know about central simple algebras until now) Since in the MSE, all automorphisms over matrix rings of fields are inner. Now $M=M_n(\Bbb{R})\oplus AM_n(\Bbb{R})$ is a subring of $M_{n\times n}(\Bbb{C})$. Since the latter has all automorphism being inner it follows it is also true for its subring (as the ideal of a subring is inherited from its parent ring and the multiplicative operation of the subring $M$ is non zero).
Therefore for any two automorphisms $f,g : M_1\otimes I + M_2 \otimes A \mapsto M_1 \otimes I - M_2 \otimes A$ they should be
I'll work out the exact form of that isomorphism later, cause typing on someone else's car is too unstable
 
6:37 AM
@Lepidopterist if $f(v)=a+bv$ then clearly $ax+by-z$ is in the kernel.
or e.g. if $f(v)=v^2$ then $xz-y^2$ is in the kernel
 
$f, g : A → B$, $g(a) = b · f(a) · b^{−1}$
 
@Secret Your $M$ is not a subset of $M_n(\Bbb C)$ unless $A=iI$. You are making complexification more complicated than it needs to be. And your claim that [all automorphisms of a ring are inner $\Rightarrow$ all automorphisms of a subring are inner] is false.
 
but complexification needs some element such that under multiplication, behaves like the imaginary unit. How can I claim that $A=iI$ when building a complexification of the vector space of real matrices?
 
6:53 AM
You can use your vocal cords or your typing fingers to make the claim
 
but what tells us that $A=iI$ is the only possibility, or we only need a possibility instead of it being unique?
 
You want a vector space over $\Bbb C$, so you need to be able to multiply by $i$, so you need to have $i$ times your original things in the new vector space
You can also get an isomorphic copy of the complexification over an isomorphic copy of $\Bbb C$ that doesn't use the letter $i$, in practice, but you don't do that unless there's a reason to
If $V$ is a real vector space, the complexification is $V\otimes_{\Bbb R}\Bbb C$, and every element of the complexification is uniquely expressible as $v_1\otimes1 + v_2\otimes i$ where $v_1,v_2\in V$.
 
ok I see, so using $i$ is sufficient
btw
> your claim that [all automorphisms of a ring are inner ⇒ all automorphisms of a subring are inner] is false.
I don't quite understand, isn't that if a ring has only the trivial ideal, then it follows its subring can only have the trivial ideal, thus as long the multiplicative operation is nonzero then it will be a central simple algebra and thus Skolem Noether theorem applies?
 
[a] $\Bbb Q$ has only the trivial ideals $(0)$ and $(1)$, but it's subring $\Bbb Z$ has nontrivial ideals. [b] how in the world is that relevant to what you quoted me saying?
 
[b] because Skolem Noether theorem implies all automorphism of a ring is inner as per this MSE: math.stackexchange.com/questions/188684/…
 
7:05 AM
Every automorphism is $M_n(k)$, yes, but what's your point? That doesn't mean subrings of $M_n(k)$ have only inner automorphisms.
For instance, there is an isomorphic copy of $\Bbb C$ within $M_2(\Bbb R)$. Now $M_2(\Bbb R)$ has only inner automorphisms but $\Bbb C$'s complex conjugation is not inner (there are no nontrivial inner automorphisms in a commutative algebra!)
of course, complex conjugation in the isomorphic copy of $\Bbb C$ can be realized as $f(A)=BAB^{-1}$ for some matrix $B$ outside of the isomorphic copy of $\Bbb C$
Couple more examples of things: the algebra $M_2(\Bbb R)$ has no nontrivial two-sided ideals, but the subring of diagonal matrices (isomorphic to $\Bbb R\oplus\Bbb R$) does have two nontrivial ideals. And the algebra of matrices generated by the identity matrix and any nonzero strictly upper triangular matrix also has a nontrivial ideal: the set of strictly upper triangular matrices.
 
[b]Ah, I should have made my statement clearer: I was using that MSE to my complexification problem plus my reasoning by observing that direct sum of $M_n(\Bbb{R})$ give block diagonal matrices that live in $M_{n\times n}(\Bbb{R})$. I then noting that $AM_n(\Bbb{R})$ (where $A \in M_n(\Bbb{C}$) is a subring of $M_n(\Bbb{C})$ (and thus $M_n\Bbb{R}\oplus AM_n\Bbb{R}$ will be a subring of $M_{n\times n}(\Bbb{C})$) which caused me to try to show the automorphism result I want to apply works for this subring.
(So one reason why sometimes my response does not seemed relevant is because after seeing a reply, there's actually a huge bunch of thinking going on under the hood before my response is typed, and they are sometimes too long to be explained)
[a] Ok I think I understood now. For example in $\Bbb{Z}$, $2\Bbb{Z}$ is a nontrivial ideal because product of even numbers are even, but in $\Bbb{Q}$ there is some $q \in \Bbb{Q}$ such that $2q$ is odd, thus $2\Bbb{Q}$ escaped from becoming an ideal
(and similar argument applies for the matrix examples you mentioned)
I guess the above conservation shows I am still not good with calculating ideals, and should work more on it
 
7:26 AM
Uh oh, when I suggested you do Dennis and Farb, perhaps I should have suggested you do some ring/field theory first :P
 
yeah, back in that self study, I never expected group theory is so massive that I end up stuck in there and never really read into rings yet
 
Assuming that's what is going on here...
 
well... er... the whole conversation starts because of a free association that occurs when I read a certain message:
4 hours ago, by Leaky Nun
6 mins ago, by Leaky Nun
well, exp(A)* exp(A) = exp(A*) exp(A) = exp(A*+A) = exp(0) = I
which somehow caused me to think about exp(A+B)=-I which then leads to the question of pondering $A^2=-I$ and then leading to trying to complexify the matrix of reals
These free associations takes less than 5 mins each, and they zip from one to the next like lightning arcs
and often that's how I spotaneously go into experiment mode in mathematics
 
7:51 AM
Anon: My above clarification in picture:
Typing all of that out then becomes a paragraph because there are so many steps in it
 
yeah, I ain't reading that
 
Otherwise I don't know how to condense my whole thought process in my response so you can see why I am actually responding relevantly to the quote
(something tangential) I am not sure if the following observation has a name: Consider the ring $\Bbb{Q}$. Note if your starting element is some rational $q$ with odd denominator, then $2^kq$ for any integer $k$ will still be an irreducible fraction, but for rationals $r$ with even denominators, eventually there is some $u$ such that $2^ur \in 2 \Bbb{Z}$. I wonder if there's a name for this distinction
 
What power of $2$ do you multiply $1/6$ by to make it an even integer?
 
sorry typo, I mean dynadic fractions $\frac{m}{2^n}$
 
sure, and if it were of the form $m/p^n$ for any prime $p$ then there exists a power of $p$ that we can multiply by to make it in $p\Bbb Z$. nothing special with $2$, nothing special at all to deserve a name, as far as I can see
 
8:06 AM
More generally, let some ring $R$. For all $r \in R$ and for some $k,\ell \in R$, there seemed to be at least two classes of elements. One class has the property that $k^nr \in S$ where $S \subset R$ and for any element $s_1,s_2 \in S$ $s_1s_2$ is also in $S$ (i.e. multiplication is closed for elements within $S$). Another class has the property where no matter how large the integer $n$ is, $\ell^nr$ does not converge to any subset of $R$ with the above closure property.
Are these classes special enough that they have a name?
The above in simple english, is that there seemed to be a class where when you iteratively multiply the elements within by a specific element, it eventually get stuck inside some subset, whereas there's another class where this never happened
 
 
1 hour later…
9:11 AM
[cont.]
Previously we explore:
$$\int^{\int^{\int}}dx$$
where $\int^{\int^{\int}} = e^{e^{\int \ln \int}\ln \int}$
Using the taylor series for ln and exp, we have:
\begin{align}e^{\int \ln \int} & = \sum_{n=0}^{\infty} \frac{(\int \sum_{m=1}^{\infty}(-1)^{m+1}\frac{(\int - 1)^m}{m})^n}{n!}\\ & = \sum_{n=0}^{\infty} \frac{(\int \sum_{m=1}^{\infty}(-1)^{m+1}\frac{\sum_{p=0}^m\binom{m}{p}(-1)^{m-p}\int^{(m)}}‌​{m})^n}{n!}\\ & = \sum_{n=0}^{\infty} \left(\int \sum_{m=1}^{\infty}\int^{(m)}\sum_{p=0}^m\frac{(-1)^{2m+1-p}\binom{m}{p}}{(n!)^{‌​\frac{1}{n}}m}\right)^n\\ \end{align}
\begin{align}
e^{e^{\int \ln \int}\ln \int} & = \sum_{s=0}^{\infty}\frac{\left(\sum_{n=0}^{\infty} \left(\int \sum_{m=1}^{\infty}\int^{(m)}\sum_{p=0}^m\frac{(-1)^{2m+1-p}\binom{m}{p}}{(n!)^{\frac{1}{n}}m}\right)^n\sum_{q=1}^{\infty}\int^{(q)}\sum_{r=0}^q\frac{(-1)^{2q+1-r}\binom{q}{r}}{q}\right)^s}{s!}\\
& = \sum_{s=0}^{\infty}\left(\sum_{q=1}^{\infty}\sum_{n=0}^{\infty} \left(\int \sum_{m=1}^{\infty}\int^{(m)}\sum_{p=0}^m\frac{(-1)^{2m+1-p}\binom{m}{p}}{(n!)^{\frac{1}{n}}m}\right)^n\int^{(q)}\sum_{r=0}^q\frac{(-1)^{2q+1-r}\binom{q}{r}}{(s!)^{\frac{1}{s}}q}\right)^s
Note the innermost summation of each term can be simplified (writing...)
$$\sum_{r=0}^q \frac{(-1)^{2q+1-r}\binom{q}{r}}{(s!)^{\frac{1}{s}}q} = \frac{1}{(s!)^{\frac{1}{s}}q}\sum_{r=0}^q (-1)^{1-r}\binom{q}{r}$$
For the latter sum, observed that if q is odd, then the sum vanishes because of the symmetry $\binom{q}{r}=\binom{q}{q-r}$
Likewise, if q is even, then we have $\binom{q}{\lfloor \frac{q}{2}\rfloor}+2\sum_{r=0}^{\lfloor \frac{q}{2}\rfloor-1}\binom{q}{r}$
typo
$\binom{q}{\lfloor \frac{q}{2}\rfloor}+2\sum_{r=0}^{\lfloor \frac{q}{2}\rfloor-1}(-1)^r\binom{q}{r}$
in particular:
$$\sum_{r=0}^{\lfloor \frac{q}{2}\rfloor-1}(-1)^r\binom{q}{r} = (-1)^{\lfloor\frac{q}{2}\rfloor-1}\binom{q-1}{\lfloor\frac{q}{2}\rfloor-1}$$
actually sorry, there's a simpler way:
Again we consider the sum 4 lines above
We then make the observation that $(-1)^{-a}=(-1)^{a}$ for any integer $a$ to rewrite it as:
$$\sum_{r=0}^q \frac{(-1)^{2q+1-r}\binom{q}{r}}{(s!)^{\frac{1}{s}}q} = \frac{1}{(s!)^{\frac{1}{s}}q}\sum_{r=0}^q (-1)^{1-r}\binom{q}{r} = \frac{1}{(s!)^{\frac{1}{s}}q}\sum_{r=0}^q (-1)^{r-1}\binom{q}{r} = -\frac{1}{(s!)^{\frac{1}{s}}q}\sum_{r=0}^q (-1)^{r}\binom{q}{r}$$
no by the result in ProofWiki, we found the sum is equal to $\delta_{q0}$
and hence the integral simplifies to:
\begin{align}
e^{e^{\int \ln \int}\ln \int}& = \sum_{s=0}^{\infty}\left(\sum_{q=1}^{\infty}\sum_{n=0}^{\infty} \left(\int \sum_{m=1}^{\infty}\int^{(m)}\sum_{p=0}^m\frac{(-1)^{2m+1-p}\binom{m}{p}}{(n!)^{\frac{1}{n}}m}\right)^n\int^{(q)}\sum_{r=0}^q\frac{(-1)^{2q+1-r}\binom{q}{r}}{(s!)^{\frac{1}{s}}q}\right)^s\\ & = \sum_{s=0}^{\infty}\left(\sum_{q=1}^{\infty}\sum_{n=0}^{\infty} \left(\int \sum_{m=1}^{\infty}\int^{(m)}\left(-\frac{\delta_{m0}}{(n!)^{\frac{1}{n}}m}\right)\right)^n\int^{(q)}\left(-\frac{\delta_{q0}}{(s!)^{\frac{1}{s}}q}\right)\right)^s
Now because both m and q are never zero, it means all the terms with dirac deltas vanishes, can be bubbled out from the integrals (since they have no dependence on x) and thus we have:
\begin{align}
e^{e^{\int \ln \int}\ln \int}& = \sum_{s=0}^{\infty}\left(\frac{1}{(s!)^{\frac{1}{s}}}\sum_{q=1}^{\infty}\left(\frac{\delta_{q0}}{q}\right)\sum_{n=0}^{\infty} C^n\int^{(q)}\right)^s\\
& = \sum_{s=0}^{\infty}\left(\frac{1}{(s!)^{\frac{1}{s}}}\sum_{q=1}^{\infty}\left(\frac{\delta_{q0}}{q}\right)\sum_{n=0}^{\infty} C^n\int^{(q)}\right)^s
\end{align}
actually typo
\begin{align}
e^{e^{\int \ln \int}\ln \int}& = \sum_{s=0}^{\infty}\left(\frac{1}{(s!)^{\frac{1}{s}}}\sum_{q=1}^{\infty}\left(\frac{\delta_{q0}}{q}\right)\sum_{n=0}^{\infty} C^n\int^{(q)}\right)^s\\
& = \sum_{s=0}^{\infty}\left(\frac{1}{(s!)^{\frac{1}{s}}}\sum_{q=1}^{\infty}\left(\frac{\delta_{q0}}{q}\right)\sum_{n=0}^{\infty} \frac{C^n}{n!}\int^{(q)}\right)^s\\
& = \sum_{s=0}^{\infty}\left(\frac{1}{(s!)^{\frac{1}{s}}}0\right)^s\\
& = 0
\end{align}
and therefore, for any integrable function $f$
$$\int^{\int^{\int}}f(x) dx = 0$$
meanwhile:
$$\int^{\int} f(x) dx = \sum_{n=0}^{\infty}\frac{C^n}{n!}=e^C-1$$
The more interesting question for future work for any interested readers is whether there exists other integral forms that has such "singular" behaviour in the space of all integrable functions. i.e.:
$$g\left(\int\right) f(x) dx = h(x)$$
where $g$ is some relation, $f$ is any integrable function and $h$ is some fixed function
 
Anonymous
10:24 AM
What does $\Bbb{R}[x]/I$ stand for? I know the $\Bbb{R}[x]$ stands for ring of polynomials with coefficients in $\Bbb{R}$
 
Anonymous
I guess it's called $\Bbb{R}[x]$ modulo $I$, but not sure what that means
 
quotient ring of $\Bbb{R}[x]$ over $I$
 
Anonymous
What is $I$ ?
 
some ideal of the ring
The quotient ring consists of equivalence classes of the elements of $\Bbb{R}[x]$ in terms of $I$ if I recall. In polynomial rings, the quotient ring is often related to factorization of polynomials if I recall
 
Anonymous
10:37 AM
@Secret I see. So this somewhat makes sense to me now. A subset $I$ of a ring $R$ is an ideal iff $I\neq\phi$ & $\forall a,b \in I, (a-b)\in I$ & $\forall a\in I, r\in R, ra\in I$. For example ring of matrices $\{\{a,0\},\{b,c\}\}$ has an ideal member $\{\{0,0\},\{k,0\}\}$ since it satisfies the above condition. I was stuck in a related question, where they're trying to find equivalence classes of the remainder, when dividing a member of $\Bbb{R}[x]$ by $(x^2+1)$.
 
yeah for that they will be elements in $\Bbb{R}[x]/<x^2+1>$
 
Anonymous
Am I right in this $ax+b$ and $a'x+b'$ belong to the same equivalent class (in $\Bbb{R}[x]/<x^2+1>$) iff $a=a'$ and $b=b'$ ($a,b,a',b'\in \Bbb{R}$) ?
 
Anonymous
If I'm not misinterpreting then $ax+b\in \Bbb{R}[x]/<x^2+1>$ ?
 
yeah if I have not mistaken, by the division algorithm, any polynomial $p(x)$ in a polynomial ring can be written as $p(x) = d(x)q(x)+r(x)$. So the ideal determines the $q(x)$ and the remainder is the $r(x)$ in the quotient ring
thus two polynomials are in the same equivalence class iff their remainders agree, which is like what you said (I think)
 
Anonymous
I'm a bit worried that checking $a=a',b=b'$ , might not be sufficient but not necessary in case they're not numbers
 
10:51 AM
well, I don't know how else they can differ, as you said, any remainder of a polynomial after dividing by $x^2+1$ must be at most a degree 1 polynomial, thus it only has 2 degrees of freedom a and b
 
Anonymous
Yeah, in this case the coefficients of polynomials in $\Bbb{R}[x]$ are numbers. But say if they were matrices or something
 
Anonymous
Hmm, I guess I could check if their difference belongs to ideal
 
Anonymous
"In general, a quotient ring is a set of equivalence classes where $[x]=[y]$ iff $x-y \in a$"
 
actually wait a sec... I am not very sure if the remainders are $ax+b$ and $acx+bc$, then whether they are in the same equivalence class
 
Anonymous
10:56 AM
So uh, our ideal ring in this case are just formed the multiples of $(x^2+1)$
 
Anonymous
Including $0$
 
Anonymous
So $ax+b$ and $acx+b$ would belong to the same equivalence class iff $(ax+b)-(acx+bc)\in I$
 
Anonymous
Which isn't true unless and until $c=1$, considering $a,b$ are non-zero.
 
Anonymous
amirite?
 
you missed a c in acx+b
 
Anonymous
10:58 AM
amiritenow? :P
 
$ax+b-acx+bc = (1-c)(ax+b)$ hmm...
ah, the ideal $<x^2+1>$ does not contain terms in x, ok then
sounds valid
 
Anonymous
Yup. That's the main point
 
Anonymous
@Secret And rings follow that cool distributive law. So you can do that
 
yeah, rings are really nice structures, and polynomial rings, numbers commute, thus I can pull out the $c$
In matrix rings, however, you might need to be careful since it is noncommutative
but then I don't know much about matrix rings
 
Anonymous
True. I too haven't dealt with matrix rings, much. I'll perhaps trying doing an example for $2\times 2$
 
Anonymous
11:04 AM
It shouldn't be too difficult using first principles. Anyway, thanks a lot for the help!
 
11:22 AM
Currently discovering that actually writing down the things I've learned over this year for my dissertation is incredibly difficult, but that once I have written it down everything becomes a lot easier to understand
Perhaps that's just because writing it down requires me to actually think a lot harder about stuff
 
That might also be because when you write everything in a few pages, you can read a lot of content at the same time in the pages, thus gives you some kind of "bird's eye view" of what you learnt so far
 
Yeah maybe, it's also because creating some kind of logical ordering to the content requires you to really think about the content fits together
I would say
 
My attempt: N=(p1p2)(p2p3)(p3p1)=(p1)^2 * (p2)^2 * (p3) and the conclusion I could draw from that is that their distinct.
 
11:47 AM
Hey guys ! I'm trying to solve $$\int\sqrt{1+2\cot x(\csc x+\cot x)}\,dx$$
After converting to $\sin$ and $\cos$ I'm not getting anything meaningful.
 
$(\cot x + \csc x)^2 = \cot^2 x + \csc^2 x + 2 \cot x \csc x$
$\csc^2 x = 1 + \cot^2 x$
I don't know if that might help
I'm out :P
 
@LeakyNun I've already seen that , but I'm looking for an approach ( more logical )
 
$(\cot x + \csc x)^2 = 1 + 2\cot^2 x + 2 \cot x \csc x = 1+2\cot x(\csc x + \cot x)$
I've already told you
 
@LeakyNun huh , thanks ! I should've got that .
 
Hey @LeakyNun can you help me with an easy "elementary" number theory problem ;)?
imgur.com/a/mA2XG
My attempt: N=(p1p2)(p2p3)(p3p1)=(p1)^2 * (p2)^2 * (p3) and the conclusion I could draw from that is that their distinct.
 
12:12 PM
sorry, I have a flight to catch in 5 hours
 
Alright see yah, Leaky.
 
How do I eliminate $t$ from these two expressions to get an expression solely in $x$ and $y$ . $$y=\dfrac{3t}{t^2-1}$$ and $$x=\dfrac{3+3t^2}{t^2-1}$$
 
I think you need to use partial fractions
3t=A(t-1)+B(t+1)
Wait are you doing implicit differentiation?
 
12:27 PM
@DarkVampiricAbstractArtist Nope , it's just a locus problem . I need to find a relation in $x$ and $y$
 
12:47 PM
\frac{-3+\sqrt{4y^2+8yx+4x^2-27}}{2\left(3-y-x\right)}
It's pretty messy.
 
Jan
1:10 PM
When the algorithm derived by majorization-minimization, do we measure the error from the majorized function (not back to the original function)?
 
1:44 PM
Hi all, can someone help me make sense of the following graph?:
I produced the graph from a collection of my own results:
Let f(n) be the product of all the nonzero digits of a number, so that p(n) is the sum of all f(n)$\le $n
S(n) is simply the sum of all consecutive integers$\le $n, basically n(n+1)/2
D(n) is the sum of all the digits of all the integers $\le $n
Now, I simply don't understand the relationship between P(n), S(n), and D(n)
Any thoughts would be greatly appreciated
 
2:07 PM
There is nothing bizarre about that, except it is inconsistent with ZF because it breaks the axiom of foundation for a set to be able to contain itself
 
hello there! in the image i can't understand how signum function and minus sign got there and transformed $dx$ into $d\frac{1}{|x|}$ (i know that $\sign (x) = $d/dx \cdot |x|$), could some body clarify this?
 
@Secret What is ZF?
 
One of the set theories
 
I don't understand what you mean; what is the axiom of the foundation, etc.? Sorry, I'm not very well-versed in this
 
2:38 PM
well, its something from the foundations of mathematics, which has nothing to do with your question in case you are wondering
Btw, I suck at number theory the most, thus I cannot answer many number theory questions
 
hmmm... now to try visualising matrices with these...
(as for the maths details of the above 3B1B video, see this: sciencedirect.com/science/article/pii/S089812211300669X)
 

« first day (2790 days earlier)      last day (2223 days later) »