Serving the Quantitative Finance Community

Search found 306 matches

  • 1
  • 3
  • 4
  • 5
  • 6
  • 7
  • 21
by sammus
May 3rd, 2005, 5:58 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

a nice numerical analysis paper just looking at its easy-to-understand equations
by sammus
April 27th, 2005, 9:58 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

Oh! Thanks for the hints. I didnt know that. I did try one-sided FDM for the degenerated PDE at the boundary but it seemed even worse. Does it mean the existing ADE methods aint no unconditionally stable for the diffusion-convection-reaction PDE? Gotta take a serious look at the M-matrix...
by sammus
April 27th, 2005, 1:37 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>QuoteOriginally posted by: CuchulainnWe now come to implementing the linearity BC (LBC) with FDM. There are 2 scenarios. In both case we use ghost points:S1: Assume that the FDM variant of Vss = 0 is true at j = J (the ‘truncated’ far field TFF)S2: assume only that Vss = 0 and then approximate Vt...
by sammus
April 26th, 2005, 8:02 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

sorry a dumb question, can we say Urr=0 given that Ur=0? > A is a M-matrix, i.e. off-diagonals elements are non-positive and diag. els. postive and A is diagronally dominant.yeah A is a bi-diagonal matrix to be exact. I appreciate your time.
by sammus
April 26th, 2005, 7:29 pm
Forum: Careers Forum
Topic: Columbia's MAFN program
Replies: 44
Views: 162443

Columbia's MAFN program

Hi my TT,If I get it right, you were admitted by Columbia MFE program correct?
by sammus
April 26th, 2005, 3:17 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>Hi Cuchulainn,Yes I've read thru the pdf about stability analysis of the linearity boundary. Looks like they are saying the "linear boundary condition" outperforms the "PDE boundary condition" (the method introduced in Tavella's book) followed by a bunch of proofs. Both of those boundary conditio...
by sammus
April 25th, 2005, 2:28 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>very interesting. Thanks for all your insightful inputs.FOr the linearity boundary condition Urr=0, I used the reduced PDE Vt + aVr + bV = 0 (either one-sided or not) to get U(J,n+1) and went back to the ghost point U(J+1,n+1) by differencing Urr=0 after sweeping back for the next time level use....
by sammus
April 23rd, 2005, 2:05 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

Is it really? that C++ is able to reduce times by 100? have a nice weekend,
by sammus
April 22nd, 2005, 2:58 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

Hey Alan!Thanks for your concern. The running times are seconds per 1 solution, and they are almost the same for same "rmax"es in each method. So I put similar signs in front. I was using Matlab/desktop. I will try them on xbox next time.
by sammus
April 21st, 2005, 4:12 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

Thanks, I will email you.
by sammus
April 21st, 2005, 2:34 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>Just like to add some progresses during these days. In summary, we have the following options:for the structures of ADE, we have A1, the original ADE which sweeps back and forth, ie. forward on odd time levels and backward on even time levels and the computation at each time level is based on the...
by sammus
April 14th, 2005, 3:43 am
Forum: Off Topic
Topic: CS paper generator
Replies: 2
Views: 152895

CS paper generator

Try on dis, then you can get your own CS paper unless you spell your name wrong
by sammus
April 12th, 2005, 3:01 am
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>My pleasure. The whole set of the FDM equations could be, in one case,Diffusion-Advection-Reaction PDE:from right to left (we use backward sweep here): that is,Perturbed boundary at r=rmax yields:, where i goes from 1 to I.After killing off the ghost point, the value at that boundary readsfrom le...
by sammus
April 11th, 2005, 5:39 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

<t>Hi yall,I am now convinced that the ADE is unconditionally stable for heat equations. I've been playing around with the ADE for the CIR bond PDE and found something interesting. Seems like the ADE is conditionally stable (hope I was comepletely wrong) for Advection-Diffusion-Reaction equations, i...
by sammus
April 7th, 2005, 4:05 pm
Forum: Numerical Methods Forum
Topic: Finite difference - CFD technique
Replies: 525
Views: 318440

Finite difference - CFD technique

QuoteOriginally posted by: ArnheimSamin each timesteps you do 2 sweeps with dt,ie in fact you dont go to T but to 2T.alternate fwd/bkwd sweep.ie j even: fwd sweepj odd: bwd sweepthat was what I did in an alternate fashion, but...I will double check it.
  • 1
  • 3
  • 4
  • 5
  • 6
  • 7
  • 21