last day (14 days later) » 

7:12 AM
Hey, you free for a while?
 
Yes, I kind of am.
 
I just wanted to ask you about what you plan to do after school.
I'm pretty sure you're not interested in KVPY.
Or JEE, for that matter.
 
What d'you mean by after school?
 
Post-12
Where are you going to apply?
 
Oh, no. I guess I will just get into a decent uni to study math.
@SohamChowdhury Haven't thought about it.
 
7:16 AM
What do you mean by "decent"?
ISI, CMI etc. all have olympiad-ish entrances.
And abroad is sort of hard without millions of extracurricular achievements and everything, I guess.
I'm really worried, tbh, @Balarka.
That's why I thought of asking you for advice, as it were.
 
I don't prefer ISI/CMI. Rather, HRI or IMSc.
Sorry, my internet is horrific.
Anyway, olympiad-ish stuff isn't really hard. You just need practice.
 
I know. How do you plan to get into HRI?
Those are grad schools.
I'm talking about undergrad.
Ping me when your internet is back, please. :)
 
7:48 AM
@SohamChowdhury I have been told that you just need to have a few papers of your own along with some recommendation letters.
 
But that's for grad.
 
@SohamChowdhury Yes, I know those are grad school. I haven't got anything planned really. I know of a professor from HRI, though. (have taken his classes)
@SohamChowdhury No, actually.
 
What are you planning for undergrad?
 
I guess they'll tell me to write a few papers in 11th grade. That actually helps.
@SohamChowdhury I haven't planned anything of anything, really.
 
Who'll tell you to write papers? Actual research papers?
Wow.
Damn, you're sort of advanced, I guess.
 
7:53 AM
The professors I have taken classes with in the uni and who has agreed to guide me so that I can learn math.
@SohamChowdhury Yeah, I guess.
I have no idea what those will be about, though. I guess they will give me topics to think about.
 
Do you have any contacts -- like people who can guide me?
I've been looking for that kind of person for a while.
 
I am not sure. I was never looking for persons who can help me in my study of math -- it was a lucky chance that one of my school professors know the head of the department of the uni I go to.
try going into your nearby universities, I guess. i have no idea, really.
but i am not thinking about admissions in unis/papers/etc right now :P what else can a person want if he has time to do math?
 
Believe me, I want to sit at home and do math all day, but sometimes real life catches up with you.
College is sort of inescapable. Amar eto chinta korte bhalo lage na.
But.
 
well, if you do well at 12-th, i don't think anything can stop you to get admitted at really good uni.
you can get into ISI, say. or CMI.
 
No. They have entrances.
 
8:05 AM
CMI doesn't even care about your HS grades.
yeah.
 
They only want >80% or something.
 
right. as long as you do well at entrances, you have no problem. and CMI is really a good uni.
 
Entrances are hard. That's the whole point.
And they have little in the way of real math.
It's so INMO-ish, the CMI entrance.
 
well, you'll have to take entrances anywhere you go.
there's no getting away it.
 
I suppose so.
Could you do the CMI entrance comfortably right now?
 
8:07 AM
i haven't tried, but IMO isn't really hard. you just need some practice to cope up with the time limit.
i have seen some of the questions -- most of them are manipulations. there's a positive proportion of them where you need some creative thinking, however.
it'd be nice if you can do a lot of number theory while you're at it, though. it helps, because most of IMO is number theory.
 
Do you know the infamous 1988 Vieta jumping problem?
"If $a, b, q = \frac{a^2 + b^2}{ab + 1}$ are integers, $q$ is a square."
 
Vieta jumping is a bad way to do it
I have my own method :)
 
Which is?
No sieves pls. :P
Is it the "family of hyperbolas" proof in Engel?
(which is essentially equivalent)
 
8:25 AM
no complicated stuff needed. with my method, you can not only prove that $q$ is a square, but also that $q$ is an integer iff $(a, b) = (t, t^3)$ or $(a, b) = (t^3, t^5 - t)$
sorry, my internet is going off and on
the first case you can obtain as follows :
$\frac{a^2 + b^2}{1 + ab} = k \Rightarrow a^2 + b^2 = k + kab \Rightarrow a^2 - (kb)a + (b^2 - k) = 0$
Since $a$ is an integer, discriminant must be a perfect square.
i.e., $b^2 k^2 - 4b^2 k^2 - 4b^2$. multiply by $b^2$ without loss of perfect square-ness : $b^4k^2 - 4b^4 k^2 - 4b^4 = (b^2k + 2)^2 - (4 + 4b^4) \leq (b^2k + 2)^2$
Case I : if $k \geq b^2$, then $4b^2k - 4b^4 \geq 0$, which implies $b^4k^2 - 4b^4k^2 - 4b^4 \geq (b^2k)^2$.
so this thing can either be $(b^2k)^2, (b^2k + 1)^2, (b^2k + 2)^2$. the last case is not possible, as $4 + 4b^4 \geq 0$. that said, the middle case is not possible as otherwise $k$ wouldn't be an integer (verify this).
hence, $4b^2k - 4b^4 = 0$, hence $k = b^2$, a perfect square. manipulation gives $b = a^3$, proving a part of our claim.
Case II : $k < b^2$. this is harder, but no more than some manipulation. you'd end up with $a$ being a cube and $b = a^{5/3} - a^{1/3}$, proving the second part of our claim.
@Soham I have typos in there. The discriminant is $k^2b^2 - 4(b^2 - k)$. but my overall calculation holds.
 
9:01 AM
Thanks, I'm going through it atm.
 

last day (14 days later) »