« first day (4714 days earlier)      last day (603 days later) » 
00:00 - 21:0021:00 - 00:00

21:00
There is a lot of notation, Samir. It is very far from my expertise; I'm lost in the first two sentences.
@SAMIRORUJOV I don not understand the second sentence. Are you saying that $P[X=b] = f(b,p)$ for some binary vector $b$?
in which case it would sum to one and be non negative, so am further lost in the first displayed equation.
And, to clarify, $\bar{\mathbf p}$ is the mean? Or just an arbitrary value?
@copper.hat X is a vector. f(X,p) is a multivariate probability mass function. So yes you are right
$\bar{\mathbf p}$ is an arbitrary value of the parameter vector of the probability mass function.
@SAMIRORUJOV I think you need to clean up the question a bit. I don't want to spend my Saturday figuring out the question.
What is $\gamma$, etc. There is a lot of odd stuff there.
Why did the indicator function disappear on the second displayed equation?
Etc, etc.
Yeah seems so. Maybe It would be better to ask this: can 3 and 4 be implied from 2a? That is all I need to know.
21:04
Ted out of curiosity, why did you decide to do Ch.7 before Ch. 6?
@SAMIRORUJOV I do not understand the setup, so leaping ahead is difficult.
@copper.hat should've been sweated out from the morning bike ride...
@D.C.theIII I rode yesterday at friendsofchinacamp.org/park-map
@copper.hat Thanks for having a look. I will try to write everything more clearly. I was just afraid that it will be too long.
Trying to decide what to do today.
@SAMIRORUJOV You should note that $f$ is differentiable as well.
21:08
@copper.hat Right!
@D.C.theIII some friends (a professor from Ireland & family) were supposed to visit for dinner tonight, but unfortunately his wife (yet another Irish professor) fell and broke her elbow. She is in surgery today.
I went for the ride on Friday so I would have time to prep for dinner (I am relegated to cleaning duties)
That's unfortunate. Hope they have travel insurance. The park looks mountainous.
The cleaning duties most likely are still going to be forced upon you though even though there is no dinner to prepare for :p
@D.C.theIII I assume so (insurance), pretty standard for Europeans visiting the USA
Do you know any online course on real analysis. I really need that. I have never had a formal math education and somehow found myself in the middle of math problems. I would appreciate if you could give me some advice on this journey.
My brother has treated many Americans in Ireland, who never pay.
21:11
Being a professor I would assume he would be responsible enough to deal with all those contingencies. Not like he is a young backpacker and has to cut corners.
@D.C.theIII Because it's more basic and chapter 6 leads into chapter 8. Our semester ended with chapter 5. If there had been more time, I might have concluded with chapter 6.
@TedShifrin Could I ask which book are you talking about?
Multivariable Mathematics by Ted Shifrin
or by Theodore Shifrin rather
It is an integrated treatment of linear algebra and multivariable calculus/analysis.
Are you the author?
21:13
There are 112 YouTube videos of the course, if you're interested.
Yes.
I retired from teaching about 8 years ago ...
Woow. This is a cool place to chat.
Low pressure place to talk and learn how to talk mathematics.
@SAMIRORUJOV I personally would have found real analysis hard to learn on my own, but that could reflect my learning style.
There were many things I thought were blatantly obvious, but they turned out to be wrong.
@SAMIRORUJOV MIT has uploaded a course on real analysis
What could I do then? Can I take non-degree courses for free at a good university?
21:16
i learned real analysis on my own, you can do it too
i did have two calculus classes formally first tho
@shintuku I also had calculus but it was for economists not for real mathematicians
@shin
I took a remedial 104 real analysis class when i started at Berkeley and found it tough. Later I looked at my high school & undergraduate math notes and found I had been exposed to exactly the same material, but clearly it had not sunk in.
Lots of the students in the Honors course I taught (based on the text) were econ majors who either double-majored in math or who wanted to do serious grad work in econ. Calculus for econ students is too watered-down a course to be of much value.
very basic stuff, $\inf,\sup$, continuity, connectedness, series, convergence, inverse/implicit function theorem stuff.
@shintuku What is your current level in real analysis? Are you teaching it somewhere?
21:19
hmmm interesting.....some interesting war stories on real analysis.
You took both semesters, @copper? The multivariable stuff certainly wasn't in 104A.
@SAMIRORUJOV no lol i'm an economics undergraduate student
never took a formal test
@TedShifrin THis is what led me down this path of masochism. Doing finance and said this is not rigourous enough for me....I wanted to know "how" theses black box formulas worked.....and now I'm here.......
@copper.hat I do know simple things about the concepts you mentioned. I can do some easy proofs, have never dealt with it seriously.
Well, for lots of people, the black boxes are more than sufficient.
21:21
@TedShifrin I found your youtube channel. Thanks for this.
I guess I have too strong of an urge for curiousity
@SAMIRORUJOV I think part of it, for me, was spending long hours trying to figure things out.
Sure, @Samir. It has all the proofs (pretty much) and also plenty of concrete examples/computations. If you're interested in homework exercises, you pretty much need the book to get those. D.C. can whine plenty about those :P
@D.C.theIII I think I am also moving on the same path. when did you switch from finance to math?
😢😢😢😢😢😢
@SAMIRORUJOV too long ago..........I was one course away from finishing my degree and decided to toss it in the bushes.
21:24
the way i did it was fully review calculus on khan academy accompanied with Stewart's calculus and then i started doing problems here and there in Thomson et. al.'s Elementary Real Analysis, and then i started mixing it up with Spivak's Calculus which i found easier in some parts
@TedShifrin ))
but if i had to do it again i would stick very closely to an online youtube course, like MIT's, doing all their exercises in the order it is presented
So if you have some strong beliefs on what you want to do I would give it serious consideration....it is not for the light hearted.
The expression is ... It's not for the faint of heart. :D
@D.C.theIII I love math. But unfortunately, I am above 30 now. It is a bit late.
21:28
ted's videos are great for the transition to multivariable real analysis
and linear algebra
I initially started with Salas, Hille, and Etgen's , then after some time ended up on SPivak's Calc, then Friedberg, Insel, and Spence's Linear Algebra *after having done a simpler Linear ALgebra course, and now Ted's book.
@SAMIRORUJOV Never too late. I'm also above 30, that's why I mentioned a strong sense of belief in what you're doing because there is going to be a lot of external "pressure" on what to do.
@TedShifrin Thank you. :)
@D.C.theIII Thanks for the support. I got so happy to find myself here.
Do you know any online reading groups on math?
mathSE heheh
exactly
wow, hille had a calculus book?
21:32
yea, Calculs One Variable
@shintuku are there reading groups in mathse? How could I join to intro level real analysis for example?
unless you did Spivak in your first year Calc it was usually the "step below" Spivak
that's wild, i didn't know that
@SAMIRORUJOV no not in that sense, but i used to hang out in this chatroom and ask questions whenever something didn't make sense, and if the question was more involved i would post it on the main website
@leslietownes I take it you know them?
21:34
@shintuku cool.
yeah and people love to answer real analysis questions because it is a good chance to review basic concepts so you'll find a lot of help in that sense
dc not personally :D he died a long time ago but i have some of his other books and he is a direct mathematical ancestor
@SAMIRORUJOV What Shin suggested is how you would have to do it. You're going to have to be very self motivated because it is a long journey and won't be easy. But we are blessed with the interconnectivity of the internet and folks that are altruistic enough to help.....something that I am still captivated by
Altruistic sometimes; mean other times ...
@leslie Salas & Hille ... pretty well-known. Definitely a bit superior to the usual Stewart nonsense.
But I've never taught out of it, so I can't make specific remarks.
It's not as advanced as Spivak, it is the "middle ground: text.
21:38
No, nor as advanced as Apostol.
I put Spivak above Apostol, although Apostol has multivariable and other stuff in volume 2.
@D.C.theIII how much time do you spend on math per week?
Never seen Apostol, but I'm assuming it is on even footing as Spivak's?
Is calculus to be distinguished from real analysis?
@s.harp german detected
Apostol is very carefully written, but not nearly so much depth of analysis as in Spivak, and the proof exercises are nothing comparable.
Apostol does, however, do integration first. I think he did that because the students he taught at Cal Tech who had AP credit came in thinking they knew everything. ... And also because it's historically accurate and pedagogically sound.
21:40
@SAMIRORUJOV per day: 7 - 10 hrs, some are very productive some hrs are not. And of course not every day ends up being 7-10 hrs, but that's what I aim for.
My favourite analysis book is the one by Dieundonné, but I admit I only ever read like 3 chapters of it (but the chapter on differentiation is incredible)
All 4+ volumes, s.harp?
no, the stand alone thing
I don't think there's anything stand-alone. But you mean volume 1, I presume.
I have Dieudonné in French.
It is beyond terse, but has all sorts of good stuff in it, and very incisive exercises throughout.
I never read the other volumes, I kind of want to
21:42
@TedShifrin Everybody thinks they know everything until they get introduced to Mr .$\delta$ and Ms. $\epsilon$
Later volumes cover integration theory, differentiable manifolds, some differential and Riemannian geometry, some Lie groups, even a bit of differential systems.
@D.C.theIII I am so bad with time management and not organized at all. Sometimes, I work a lot, sometimes I dont.
No, DC, not about that. It's about using the FTC as the definition of the integral without even knowing what the integral is.
I could imagine that happening. Because integration at that level is just mechanical
when first taught
Anyhow, Archimedes had the idea of the integral long before Newton came along.
21:45
ted: what was he like in person?
Lol
what a mean man you are......🤣
who is the guy on this cover?
groucho on a bad hair day
should have a warning on the cover: "WARNING this is really a bourbaki book"
Well, I find Dieudonné's style a bit friendlier than his style in the Bourbaki books ... but those are much more formal in nature.
honest people don't need aliases
21:51
Hmm, @s.harp, the multi-volume treatise is called Elements of Analysis.
@TedShifrin Treatise on Analysis*, but yeah the first volume was originally written separately and is usually sold with the title "foudnations of modern analysis"
we consider subgroups of $\mathbb Z^2$. consider $G_1$ generated on $(2,0), (0,3)$ and $G_2$ generated on $(2,0), (0,1)$. we consider their Z-module structure to represent the group homomorphism $G_1 \to G_2$ by a matrix. clearly $G_1 \to G_2$ is represented by a matrix with integer entries, but shouldn't there be a matrix with integer entries representing $G_2 \to G_1$?
I think he wrote it first and then the other volumes came later. I bought all four in Paris in 1980, I think.
the matrix for $G_1 \to G_2$ is $\begin{bmatrix} 1 & 0 \\ 0 & 3 \end{bmatrix}$ but its inverse has rational entries
there are 9 voliumes now, with more being originally planned
21:53
9???
He died 30 years ago ....
>Dieudonné is putting together a monumental treatise which will comprise d volumes; the best estimate we know on d is 8 < d < 12, but the upper limit is not certain. So far, 6 volumes have appeared in English and 2 others in French that have not yet been translated.
projecteuclid.org/journals/… (I dont know how to lin kthis
Hmm, I guess I knew that $n$ was greater than $4$, but I've never seen the later volumes.
some money grubbing corporation bought the IP and will be releasing things in the dieudonne extended universe forever
Ah, lots of harmonic analysis for several volumes.
There's a modular arithmetic (cases) formula for partial sums of: $-1 + 4 - 6 + 8 - 9 + 10 - 12 + 14 - 15 + 16 - \dots$ (composite numbers)
21:57
The last volume was differential topology. Hmm, I would like to see that one.
no inverse no isomorphism. cake but no eating
@s.harp With tricky URLs it's a good idea to wrap them, using [text](url) syntax, eg: Review: Jean Dieudonné, Treatise on analysis
22:19
So I tried a little bit of alcohol last night and…. No, just no. My body simply can’t handle it.
first time?
I tried to clean up the setup. Could you please have a look at it? I just need to make sure that the leap from Eq.(2a) to Eq.(3) and Eq.(4) is correct.
0
Q: Maximizing the weighted sum of log likelihood function?

SAMIR ORUJOVSuppose $\mathbf{X}$ is a random binary vector over the sample space $\Omega$ such that $\mathbf{X} \sim Bernoulli(\bar{\mathbf{p}})$. As a matter of notation, suppose $f(\cdot;\bar{\mathbf{p}})$ is the joint probability mass function (pmf) of $\mathbf{X}$. By definition, such a pmf is continuous...

@冥王Hades Try some math problems instead.
@shintuku Can you have a quick look at it?
i suck at probability
It is not probability.
It is just a maximization problem.
Anyone have an answer/feedback to question I asked in chat at 9:45am?
22:31
@shintuku You have a maximization of a multivariate function with the foc given in (Eq.2a). Then I deduced eq.3 and eq.4. Is this deduction correct or not. That is all
@shintuku We all suck at something until we give a try
6
@geocalc33 can you repost please?
Maybe do it over a fixed domain as an example such as $\mathbb{R}^3$ and see how the Langragian set up works out then you can generalize? Because the way it is typed up there is a lot of compact notation, which appears elegant, but if you're not in the field will appear confusing.
@D.C.theIII Great idea! Thanks.
*Does there exist a pair of smooth manifolds $(M,N),$ $M\ne N$ related by an isometry $g,$ s.t. every smooth regular foliation of $M$ satisfies some differential equation on $N,$ and every smooth regular foliation of $N$ satisfies some differential equation on $M?$*

Essentially one is taking all smooth regular foliations on $M,$ restricting to the metric of $N$ which can be obtained via the isometry $g,$ and checking whether these foliations satisfy differential equations. The same process is done with foliations on $N.$
foliational reciprocity
looking for feedback
 
1 hour later…
23:54
@leslietownes tell us about it ;-p
@TedShifrin I love dem books. Definitely my personal favorite approach to the material.
@D.C.theIII is someone trying to unseat me?
@robjohn Always. Your seat on the throne is never safe. You must defend it.
@XanderHenderson there you go, bringing politics into it again.
@XanderHenderson uneasy sits the crown
@robjohn "Poly", from the Greek for many, and "tic", English for "blood sucking parasite".
00:00 - 21:0021:00 - 00:00

« first day (4714 days earlier)      last day (603 days later) »