« first day (1680 days earlier)      last day (3639 days later) » 

21:00
@iwriteonpeoplewhowriteonbanana restart it ?
@TedShifrin Be happy, I peaked at 98°C again today
@evinda The ratio of the terms in the series is $\left|\frac{a_{n+1}}{a_n}x\right|$. If we know the limit of $\left|\frac{a_{n+1}}{a_n}\right|$, what would that tell us?
you're soon to explode, Hippa.
21:07
@TedShifrin I bet you're happy ^^
only implicitly
and only in a neighborhood where's he determined not to be nothing
@robjohn If it tends to a real number, then $\left|\frac{a_{n+1}}{a_n}x\right| \to |x|$, otherwise it doesn't depend on x, right?
LOL @DavidW
I've always found it amusing that topology explains why I can't make my hair look good.
21:12
somehow I don't think it's Euler characteristic's fault that we aren't handsome and dapper
@TedShifrin Alex was supposed to come down to my office after lunch. >:(!
I'm sure he had people telling him things he had to do, @Mike.
You can't be the center of everyone's world.
Yes, it seems likely.
Sure I can, I just have to work harder.
I'm not sure it's worth hassling with LA traffic to visit you and Jacob ... :D
Probably not.
21:17
But I do want to go to a few LA restaurants ...
Probably not worth it.
:P
and I have to meet @robjohn
I might be going up to the Bay Area a couple of times a year ... we'll see ...
@evinda why would it tend to that?
one of my favorite students from my grad school days at Berkeley teaches at CSSB ... I need to visit him, too.
@TedShifrin Clearly, to remedy the situation, I should drill a hole in my head. On second thought...
21:22
@DavidW ... Vanity is just not worth it :P
Of course, eventually balding will take care of it. I try to be an optimist, but my attempts are feeble.
LOL, @DavidW ... it may be a lost cause.
21:39
My favorite kind.
@robjohn I meant: f it tends to a real number $l \in \mathbb{R}$, then $\left|\frac{a_{n+1}}{a_n}x\right| \to |l x|$, otherwise it doesn't depend on x, right?
@evinda what does it mean if $|lx|\lt1$ or if $|lx|=1$ or if $|lx|\gt1$
That the series $\sum_{n=0}^{\infty} a_n x^n$ converges only if $|lx|<1$? @robjohn
We have the differential equation $y''+ \frac{p}{x} y'+ \frac{q}{x^2}y=0, x>0$ and we set $z=\log x$.

Then $y'=\frac{dy}{dx}=\frac{dy}{dz} \frac{dz}{dx}=\frac{1}{x} \frac{dy}{dz}$

$y''=\frac{d^2y}{dx^2}=\frac{d}{dx}\left( \frac{dy}{dx} \right)=\frac{d}{dx}\left( \frac{1}{x} \frac{dy}{dz} \right)=-\frac{1}{x^2}\frac{dy}{dx}+\frac{1}{x} \frac{d}{dx}\left( \frac{dy}{dz}\right)$

How can we find $\frac{d}{dx}\left( \frac{dy}{dx}\right)$?
@evinda it converges if $|lx|<1$ it might converge or diverge if $|lx|=1$.
what do you suggest for complex numbers for a beginner?
21:52
@user4275686 $i$
i can't imagine, @robjohn.
$1+i$ is also good
@robjohn And if it always hold that $|lx|<1$ then the radius of convergence is $+\infty$. If $|lx| \geq 1$ then can we determine something about the radius of convergence?
yeah, I got that!
You probably need to be a bit more specific, @user4275686
21:53
@user4275686 sorry, I couldn't resist :-)
no, @robjohn, I is amperage, not resistance.
you people...
you're just jealous, @Mike
@TedShifrin give it some time and you'll get a charge.
@robjohn, I am revolted.
and don't say ... ohm ...
21:55
Oh, ok! It has to be for a fourth semester engineering course. Topics like derivative,integrals of complex numbers.. Cauchy's formula/theorem. Residue theorem. contour integrals.
@TedShifrin You're impeding our conversation here.
ohhh, complex variables ...
get a book like Churchill or whatever book you're using,. @user4275686
@robjohn: you've exceeded my capacitance.
@TedShifrin i have no idea what you just said!
It's not complex numbers, @user4275686. It's calculus of complex functions. Presumably you have a text for your course?
We have the differential equation $y''+ \frac{p}{x} y'+ \frac{q}{x^2}y=0, x>0$ and we set $z=\log x$.

Then $y'=\frac{dy}{dx}=\frac{dy}{dz} \frac{dz}{dx}=\frac{1}{x} \frac{dy}{dz}$

$y''=\frac{d^2y}{dx^2}=\frac{d}{dx}\left( \frac{dy}{dx} \right)=\frac{d}{dx}\left( \frac{1}{x} \frac{dy}{dz} \right)=-\frac{1}{x^2}\frac{dy}{dx}+\frac{1}{x} \frac{d}{dx}\left( \frac{dy}{dz}\right)$

How can we find $\frac{d}{dx}\left( \frac{dy}{dx}\right)$?
@DanielFischer @quid Do you maybe have an idea?
21:57
@TedShifrin no ,not yet! profs. are being very blunt recently!
hmm, I misspelled capacitance ...
not so recently
@TedShifrin hmm... it doesn't appear so :-)
what kind of answer do you want?
LOL, I had it right, @robjohn, and spellcheck said it was wrong (because it was ignorant), so I changed it, and then time ran out. :( I hate the time limit in here.
@TedShifrin it is correct... what do you mean?
@user4275686: Google Churchill Complex Variables.
21:59
@TedShifrin I need references , texts, links, discussions.
@TedShifrin perhaps you need to refresh your page.
I changed it from a to e, and I presume you changed it back, @robjohn. Very crafty.
@user4275686: You need ...
Do some searching, then.
@TedShifrin You never had it right... I had to change an 'i' to an 'a' as well.
Oh, that was a total mistype, @robjohn. Definitely time for me to retire. :P
Wait, you're an english prof?
22:01
@DavidWheeler he's a prof that speaks English
I believe math profs should be literate, @DavidW ... and, btw, I was a French major in college :P
@TedShifrin i am. but there is so much that it starts feeling overwhelming particularly everyone has their own favourite and there are so many!
but we cannot know what fits your style, @user4275686. It's like asking us to tell you what to eat for dinner.
@robjohn but he's not actually from England? The plot thickens.
no, @DavidW, I'm a commie from Russia.
22:02
@DavidWheeler I cannot say...
@TedShifrin why do you suggest churchill? I can take it from here!
@DavidWheeler with a bit of cornstarch
@robjohn Cornstarch....yum.
I made chicken-friend steak for the first time in my life last night.
Churchill is a good book. It still has proofs, so it's more for math people, but it's accessible.
@DavidWheeler what is the ratio of chicken to friend?
22:04
LOL, @robjohn is quite the eagle eye today
he needs to get on the road and do some work :P
@user4275686: I used a book by Wunsch when I taught the undergraduate course. It has some cool engineering applications in it.
@Mike: still no Alex?
yes, that's one, @user4275686. Now find Wunsch.
He's now meeting people, @Ted. The number theory friend he wanted to meet is going to leave without meeting Alex; my friend wants to go home. :P
it's only 3 PM; good grief, @Mike.
Well, I think the next time Alex is free is 7.
22:09
LOL, oh. He'll be asleep by then.
Still, your friend should make allowances. You adults should be magnanimous.
He has a qual to study for.
so he has plenty of time to study while he's waiting?
when I was your age, I was accommodating
Is it really so hard to believe that he studies better at home? :P
I suspect he can find other number theory students to talk to.
@evinda sorry I am not sure what is meant. (and am about to leave)
22:14
@quid A ok.. See you :)
See you!
hi/bye @quid
bye @TedShifrin
Later pal
@robjohn I missed this one
8
A: How to solve a hard integral?

ChappersOkay, I'm going to do this the "special functions and pray" way: the first thing to do is produce a simpler integral; the one I'm going to deal with is $$ I(a,s) = \int_0^{\infty} \frac{x^{s-1}}{1+x^2} \arctan{\sqrt{a} x} \, dx. $$ (Another option is using $\arctan{(a+x)}$, but that looked even w...

Pretty easy, indeed.
22:17
bleh-I meant FRIED
I think @robjohn knew that, David. He's carping :P Fried carping.
That's what do you get when you cross a mean square with an eagle eye?
We have the differential equation $y''+ \frac{p}{x} y'+ \frac{q}{x^2}y=0, x>0$ and we set $z=\log x$.

Then $y'=\frac{dy}{dx}=\frac{dy}{dz} \frac{dz}{dx}=\frac{1}{x} \frac{dy}{dz}$

$y''=\frac{d^2y}{dx^2}=\frac{d}{dx}\left( \frac{dy}{dx} \right)=\frac{d}{dx}\left( \frac{1}{x} \frac{dy}{dz} \right)=-\frac{1}{x^2}\frac{dy}{dx}+\frac{1}{x} \frac{d}{dx}\left( \frac{dy}{dz}\right)$

How can we find $\frac{d}{dx}\left( \frac{dy}{dx}\right)$?
* $\frac{d}{dx}\left( \frac{dy}{dz}\right)$
@TedShifrin no hopes to find wunsch!
@evinda $\frac{d}{dx} = \frac{dz}{dx}\frac{d}{dz}$
22:34
@AntonioVargas I see.. Thanks a lot :)
@evinda np :)
@AntonioVargas Then we find that the solutions are $y_1(z)=e^{r_1 x}, y_2(z)=e^{r_2 x}$ and we have that $z=\log x$. How do we deduce that $y_1(x)=x^{r_1}, y_2(x)=x^{r_2}$?
I thought that it would be as follows:
$y_1(z)=e^{r_1 z} \Rightarrow y_1(\log x)=e^{r_1 \log x} \Rightarrow y_1(\log{e^x})=e^{r_1 \log{e^x}} \Rightarrow y_1(x)=e^{r_1 x}$.
Where am I wrong?
22:58
I wish I could enable ChatJax on my phone
Strangely, my phone has ChatJax in its dictionary so it autocompletes it for me
Hmmm
Given that I know that $f'(x) = f(x)$ and that $f(0) = 1$, how do I prove that
$$ f(1) = e = \lim_{n \to \infty} \left( 1 + \frac{1}{n}\right)^n $$?
I'm back.
23:13
mr eyeglasses, probably because you've typed it a number of times.
hi @Stan
@N3buchadnezzar Try calculating $f'(1) = f(1)$ from the limit definition of $f'$?
Hello Ted.
@n3B: Because you solve the differential equation and get $e^x$? You can check the limit by taking logs and limits.
Ah, @DavidW's suggestion is superior.
hi Jasper
hi @Clarinet
Hi @Ted
So question of the day:
hey @TedShifrin!
23:15
Dare I consider applying for a job that I know I will like at least 5x more than my current one with the possible risk that my current employer might be contacted?
lol
how are you ted? I'm trying to figure out exterior algebras :p
@Clarinetist How much money do you have in the bank?
I'm not a good person to ask such things, @Clarinet ... My answer would be that you should aim to be as happy as possible.
You're wedged between a rock and a hard place, @Stan?
Dang, not enough :P \$2k, but I'm probably going to accumulate another \$3-4k come next week
@TedShifrin LOL
23:17
My understanding is that in the "real world," people change jobs quite a bit, @Clarinet, so I don't know that your employer expects you to be there for 80 years.
Exactly, story of my math life.
yeah, @Stan, but this time it's all about wedge :P
@TedShifrin - Most people retire here with 20-30 yrs. experience. It's absurd
@DavidWheeler I tried that $$ \lim_{h \to 0} \frac{\exp(1)\exp(h) - \exp(1)}{h} = \exp(1) $$
My students who get into the computer/business world seem to be quite fluid in their jobs, @Clarinet, unlike most of us in academia.
23:18
@TedShifrin Yes, I really want to learn higher dimensional forms of curl.
And I am jobless. =(
Okay, I'm just gonna do it. And hope that my manager doesn't get pissed.
@TedShifrin You mean fluid at math i take it?
There is no curl beyond $\Bbb R^3$, @Stan, but there is $d$.
no, I meant transferring jobs with alacrity.
Anyone got time to assist me with an integral again?
I've mostly solved it
23:19
@Clarinetist If you want to apply for a new job, always go ahead with it. Whether he contacts him is not relevant.
Jasper: Be careful giving advice when you know not of what you speak.
I swear, actuaries have this thing about not pursuing other fields and being scathing onto those who do.
@TedShifrin Sorry.
The original problem was: $\int p^5ln(p)* dp$
So I've dropped differential geometry and will take real analysis, topology, and algebra next semester
23:20
@TedShifrin Not advice, just my opinion. =)
@Clarinetist: I've had former students who tried being actuaries and hated it and moved on .. others who loved it.
@Owatch do le parts
@Clarinetist You can always request that your prospective employer not contact your current employer. If you do so, I would be sure to bring some stellar examples of your work, and perhaps a recommendation from someone at your current employer you can trust not to tattle.
Well, Jasper, you stated it as factual.
And I've gotten it to: $\int p^5ln(p)* dp = p^6(lnp-1)+\frac{5}{6}p^6-5\int p^5ln(p)*dp$
23:21
Well, everything I say is my opinion, even 1+1=2. =)
Excellent advice, @DavidW ... how many layers of supervisor are there, @Clarinetist?
@TedShifrin - I've lost a lot of motivation for staying in this job long term since all I've done for the last two months is create rectangles and arrows in Word.
I've noticed that the original integral is equal to that last piece, except that it is multiplied by 5.
@TedShifrin - What do you mean? xp
So, do the algebra and solve for the integral, @Owatch.
23:23
@TedShifrin I want to move it over to the left side.
So I double the original equation, then divide by 2.
Is there someone who isn't your ultimate boss (who you fear would take reprisals) who could write for you? @Clarinet
@Ted There isn't a curl past R^3? What is $d$?
But I can't with that 5 there.
@Owatch: double??
$\int p^5*lnp*dp$
23:24
Morning Ted.
@TedShifrin "Unfortunately not" comes first to mind, but I'll have to think about that
@ted I guess I don't know how vicious the real world is.
You didn't do the integration by parts right, did you, @Owatch?
I think I did.
I don't agree with your $uv$ part, @Owatch. Nor the coefficient on your integral. You're differentiating $\ln(p)$ and integrating $p^5$, right?
23:25
@TedShifrin I thought there was a higher analog. I must have misunderstood. That's surprising lol, that was one of the things I was most hoping to learn so I could move beyond this wretched pseudovector stuff!
I'm going to call that employer tomorrow, see if they can confirm to me that my employer isn't going to get a call from them...
@TedShifrin No, I decided to do it the other way round.
Left them a voicemail today
Should I restart? Would it be better?
I know I don't have to. But if it's easier that way I will.
Curl is special to three dimensions. In all dimensions we have the exterior derivative, @Stan. What's also special to three dimensions is that we can interpret vector fields as either $1$-forms or $2$-forms. In $\Bbb R^n$, it'll be either $1$-forms (for work) or $(n-1)$-forms (for flux).
23:27
My impression is that academia tends to be a lot more lax about talking about switching jobs...
@TedShifrin?
Yes, @Clarinetist, but for the most part, except for big shots, we don't do it much.
It's easier the other way, @Owatch. But if you did your way correctly, you can add the $5\int ...$ to the other side.
I can?
Ahha, so I will get $6\int$?
Isn't that one of the rules of algebra? If $a=b-c$, then $a+c=b$?
I am going to bed, good night @ted.
g'night, Jasper. Sweet dreams.
23:28
See you later @ABeautifulMind
It looks different, so I'm scared to touch it.
I'll do that.
@Stan: If you take a vector field in $\Bbb R^4$, what passes for curl of it is a bivector with $\binom42 = 6$ components, not a vector field.
Solved
currently waiting on graduate school decision Any day now...
Try it the other way, too @Owatch. In general, you want to make the problem simpler, not harder, when you do integration by parts. With things like $\int e^x\cos x\,dx$ you have no choice but to make it comparable.
Maybe you shouldn't be applying for a new job until you know the grad school situation, @Clarinetist ...
23:33
@TedShifrin - It's an online M.S. program. Surprisingly a decent one as well...
oh, I see ... so you'd keep the job, then?
Yeah, but I would like to get out of actuarial by June
i.e., look for a non-actuarial position
Is the job you applied for non-actuarial?
Yep, every job I've applied to thus far is non-actuarial
I'll think good thoughts for you :)
23:35
Thanks @Ted
@TedShifrin Wait, I thought a bivector meant (2,0). that looks like a rank 6 tensor...
rank 6? No ... misunderstanding of rank. :)
I am allowed to split: $\int A - B$ into $\int A - \int B$ right?
I sure hope so, @Owatch.
23:39
@Ted Oh no, really? I thought the rank was the sum of the indices
Darn!
right, so it's a rank 2 tensor.
A vector is a rank 1 tensor, but a vector in $\Bbb R^n$ has $n$ components.
Oh, I always get that mixed up.
Does a bivector have $2n$ components?
or just $n$?
LOL, no, it has $\binom n2$ components.
In general, a rank 2 tensor has $n^2$ components, but if it's symmetric or skew-symmetric, that cuts down on the number of (independent) components.
I need to learn tensor and exterior products
Just to verify: $\int p^5*ln(p)*dp = ... -5*\int p^5*ln(P)*dp - ...$
23:44
:: facepalm :: What do the bottom and top stand for lol? I think I need it broken down in detail. Oh! I learned that skew-symmetric is another word for antisymmetric right? Also, this doesn't have anything to do with why John Lee talks about Christoffel symbols as $n^3$ functions does it?
Not necessarily just yet, mr eyeglasses
@TedShifrin ^
Again, @Owatch: take your answer and differentiate it to check.
Okay
Christoffel symbols are not tensors, @Stan, but they have three indices, so you count like for a tensor of rank 3, hence $n^3$.
A tensor of type $(k,\ell)$ is $k$ times covariant and $\ell$ times contravariant, so it's like a multilinear map from $V^k$ to $\otimes^\ell V$. I don't think you should be worrying about this all just yet, @Stan.
23:47
Wolf got something different for its integral.
Damnit.
Well, if your answer differentiates correctly, that means the two antiderivatives differ by a constant. That happens often.
Wolfram did it my way, not your way.
@TedShifrin I did know that Christoffel symbols aren't tensors! That's one thing I have down. I am still trying to figure out what I need to know. I'm learning a lot by stumbling around. My uncle always says doing proofs is the best way to learn.
Have you been through a solid undergraduate curves/surfaces course before you do all this esoteric stuff, @Stan?
I got $\frac{p^6*(lnp-6)}{36} + C$
What's the derivative of that? It sure doesn't look right.
23:51
I don't know, I'll check Wolfram Alpha
Do it by hand, so that you understand what you're doing.
@TedShifrin I have not! I am hoping to take one soon. I've been learning a lot on my own. I took a medical leave and that's when I started studying all this. I'm returning to classes on the 31st and I'm trying to figure out which kinds of math I will need.
Integration by parts is understanding the product rule. You'd better be able to do the product rule by hand.
How long will it take me to do it by hand?
@Stan: I feel very strongly you should go through the concrete curves/surfaces experience before you get carried away with this abstract stuff.
23:52
I have many questions to do.
You're wasting time, @Owatch.
Isn't differentiating by hand a waste of time?
I can do it, I haven't done it in a while though.
@Ted is right. Do it by hand. It will only pay off in the long run.
Or something. .
23:53
shakes head and gives up
Ok nevermind.
I'll check my notes.
(uv)' = u'v + v'u
You can't understand integration by parts if you can't do differentiation.
2
If all you care about is getting the answers, then use Wolfram on your tests and you'll be fine.
I can do it, it's just been a month or two and I can't remember off the top of my head.
fed up
23:54
Sorry @Ted
@TedShifrin Well, I trust your judgement. So what generally speaking does that mean I should stay away from? Obviously, the exterior algebra stuff seems beyond me. I also think I need a good topology and analysis course. Because one thing that has been apparent to me is that distance is not a good way of defining where you are. But to really understand coordinates independent of distance, I gather topology is used.
This is why a fourth of my differential geometry students can't do calculus ...
Oh my :/
@Ted Do you blame your school for not having better professors for calc 1 and 2?
Yes, @Stan, having a notion of distance/angles is what differential geometry imposes; having a notion of nearness is what topology does.
23:55
@StanShunpike distance is a good way of telling where you are-but not all spatial information can be squeezed to get a real number out of it
#234 in the things that bug me about actuaries: despite people thinking that they are mathematically inclined, most actuaries can't stand doing integration by parts. They memorize formulas to get around those...
I blame a lot of my colleagues for giving C's to students in calculus and intermediate courses just to get the students out of their hair, when the students really can't do anything right, mr eyeglasses. A gentleman's C in every course is bad news.
$\Bbb R^n$ with our usual understanding of it, is nice, but sometimes too special
2
@DavidWheeler Ooo that's a nice way of phrasing that. I never thought of it that way.
@David made my statement much more meaningful; thank you.
23:58
@Ted I think giving 'A's to students that can't do anything right is worse..I got 'A's for those courses yet didn't learn anything, so I can't re-take the courses to learn them better since you can only re-take courses for an improved grade and I already got the highest grade possible :(
@TedShifrin And then, your approach to wedge products using determinants made me think I should look more carefully at the differential geometry course selection and make sure I'm shooting for something that connects better with stuff I am familiar with. For instance, I am reading Tensor Analysis on Manifolds. And while I am getting more than nothing out of it, I definitely don't feel like I am soaking stuff in.
Back when I worked for an actuarial study materials company, I went out of my way to create questions which involved integrating absolute values of functions and integration by parts.
mr eyeglasses, I have spent most of my career being able to motivate most students to work their butts off and learn a lot. I am no longer able to do so with a good number of students. So I do not know whom to blame. Society has changed. We have raised kids to feel very entitled and parents make huge fusses if their kids are given bad grades that they deserve.
the determinant is THE proto-typical multi-linear functional. true story.
@TedShifrin Wow, that is sad :/
23:59
@Stan: Look at my undergraduate diff geo notes instead. There's plenty of challenge in there if you work exercises.

« first day (1680 days earlier)      last day (3639 days later) »