Thanks a lot for your valuable time and expertise. I got what I needed.

<t>Sorry if I did not make things clear enough. Let me better explain what I wanted to ask about. I work as a programmer for a credit card transaction processing company, now. I work long hours at my present job as well, when needed, until things are done. I usually never ask about money or working ...

<t>What's a problem in asking about the working hours? Of course if I have a job, then I want a better job.I need to know about the work conditions.Now, I work 9 to 5 and of course I have other things to dobeside my work duties.If I need to be there 24/7, then, well there are other jobs around And y...

<t>Tomorrow I have a job interview – research analyst – automatic trading - at a NY Hedge Fund ( also a provider of automated trading platforms). My recruiter told me to avoid negotiating salary by all means, and to say that I am looking to enter the field for a long term perspective. My background ...

Please share your ideas about how many quant jobs have been outsourced up to this time. Is this a current trend?If, yes: when, where and what kind of jobs? Many thanks.

<t>Hi. I was not able to participate in the discussion for some time.I have the following question about the solution U(t) = exp(-At)U(0).Since the discretization of the derivatives Uxx and Uyy in the heat equationcan be done only to some finite order (of accuracy, say n) in x and y, the solutionU(t...

Oops! I made a mistake. I should have been delta x squared and delta y squared in the denominators.

QuoteAn example: if we discretize the heat equation Ut= Uxx + Uyy in space only we get an ODEYou are using the method of lines here.Sorry again, butdo you keep your t (time-step) less than ?If you do, I have another question. What kind of precision are you looking at?

Sorry for my ignorance, but you made me curious.If the coefficients of your matrix are not numbers, what are they?On the other hand, if it is a theoretical discussion rather than a practical one, I'd better shut up =)

£23,000 sounds like a reasonable price

The averaged outcome (over all outcomes) of a random walk (+ drift if you want)is a solution of a second order PDE. The random walk produces diffusion in the averaged outcome. Laplasian takes care about diffusion in your PDE.

You may always do F=G(x,y)*H(t). You have to make sure though that you do not losesolutions by doing that. To be valid you need a uniquiness theorem for your problem.Hope this helps

See chapter One ofTikhonov, A. N Partial differential equations of mathematical physics [by] A. N. Tychonov and A. A. Samarski. Translated by S. Radding San Francisco, Holden-Day, 1964

Convert your matrix into Jordan normal form. Take exponent. Convert back

Reduce to heat equation and then use Laplacian in whatewer coordinate system you need.