@ACuriousMind I once spent two weeks trying to derive a result that's been bouncing around astronomy for 80 years or so. The literature search kept dead-ending at citations to papers that never mentioned the result (just people citing their friends it seems).
user54412
Unfortunately, I found that the result as quoted was exact (the prof and I suspected it might just be an approximation). We concluded that our findings were not publishable, since the response would be "well we've known that for 80 years."
@Qmechanic and other mods: There have been some suggested edits by a user named skyler , based on adding the freshly created tag "half-life" to older questions. What is the way forward?
As a reviewer, should I approve or reject that suggestion.
Currently, there is only one existing question which has that tag.
And this guy seems to be on a add-tag spree.
News flash - One of them just got approved by Brandon Enright and Frank H
It'd be worth thinking about whether we want a half-life tag. Generally I prefer not to interfere with tags, unless there is a compelling reason to do so, but I don't particularly care either way.
half-life → radioactivity
The new tag half-life introduced is really a subset, and radioactivity is sufficient for identification purposes and subsuming.
Motivation
Consider a BV function $f$ with jumps, the Fourier partial integral function (analogous to Fourier partial sum in periodic case) as $\omega\to\infty$ does converge to $f$ pointwise except possibly at jumps. But the total variation of the partial integral function does not converge to ...
@ChrisWhite Ahhh, the Ebony Warrior has the "Reflect Blows" perk. I was suddenly dying because I hit myself with a power attack, which was much stronger than any of his attacks.
user54412
1:50 PM
@RajeshD There's a lot there to digest. My one reaction is wholehearted support for wavelets. There are all sorts of bases for L^2 function spaces, and the Fourier basis is often the worst choice for the problem at hand.
@RajeshD In my opinion, appropriate window functions can bring you quite far. Making them flat at the origin gives you high asymptotic convergence rates, and making them go smoothly to zero reduces the ringing by smoothing it out appropriately.
One canonical family of window functions are the convolutions of the unit characterisitc functions with itself. The other window I use is the Hann window. I know for my applications when the Hann window is appropriate, and when the other family is better. I normally test also with truncated Gaussians and Kaiser-Bessel windows, but for the applications I had so far, the Hann window was always superior.
@vzn Yep! just for a quick project. I'm using Krauth Statistical Mechanics Algorithms and Computations, as well as a few misc papers which don't cover some things
unfortunately my algorithm is slow-ish and produces the right critical temperature only for small $N$, then diverges. So something is horribly wrong :(
From the notes I was referring to: "For one thing, we’ve seen that many classical systems are related to the same quantum system, which does not care about the lattice spacing in time. There is a set of physical quantities which agree between these different classical systems, called universal, which is the information in the quantum system."
Idk. It does seem like a fun topic of study but my thermodynamics isn't good enough for most of the serious reading material on it!
if you ask me the ising (temperature/ energy) phase transition may be some physical manifestation of the P/ NP boundary. (but probably few would understand/ agree with that statement either from physics or TCS.)
@NeuroFuzzy you seem quite adept/ advanced (at young age) but agreed physics phds sound extremely rigorous/ competitive. there is prob some variation among schools though. ivy league being most extreme. others maybe being more manageable.
@vzn Krauth also has a course on Coursera based on the book. I remember seeing him live in some conference years ago. I think his own research (at the time) was on developing new Monte Carlo algorithms (applied to looking for the glass transition or something along those lines).
@ChrisWhite Just flag it as not constructive or rude/offensive. I've never seen CuriousOne back down in comments, the only thing responding will produce there is something more rude. :P
@vzn The convolutions of the unit characteristic function with itself are just used as lowpass-filters, but I think you should be able to construct equivalent wavelets for them. I use them for sampling polygonal domain on a grid without less of accuracy.
The Hann window is used to get a frequency response from an impulse response in time. This has something to do with complex analysis, and poles caused by resonances. If you understand the resonances a bit, then you can use the Hann window (or a similar window) for suppressing their effect.
I don't think that this use of the Hann window is related to wavelets. It's rather related to chebychev acceleration, which itself is a form of poor-man's Krylov subspace method. The analog of Krylov subspace method in this context would probably be something like Prony's method.
@0celo7 If we only did category stuff...we're computing almost all things explicitly because the lecturer is old and not really comfortable with category theory. Also, because he's a number theorist and not really into the abstraction, I think.
We start with a scalar equation, everything was scalar, then we stipulate commutation relations, and those scalar functions magically turn into operators
I perfectly see the reason why one replaces momentum etc... by operators in QM, there is a perfect explanation for that, but not for replacing wave functions as operators, that makes no sense to me, our whole theory started as a theory of scalar wave functions, the only difference is relativity and magically those scalar wave functions turn into operators, do you not see the inconsistency of a scalar function turning into an operator as v ---> c?
Like, it's literally like functions deform into operators as v ---> c the way things are built right?
@bolbteppa You are not starting with a theory of wavefunctions. You are starting with an infinite-dimensional classical phase space (the field space) and essentially trying to carry out canonical quantization as you did in the finite-dimensional QM case.
Known JavaScript injected Trojan named Small.Q (probably not installed by the website's owner, but injected by attackers), I advise against visiting that link.