Serving the Quantitative Finance Community

 
User avatar
bono06
Topic Author
Posts: 0
Joined: April 20th, 2006, 11:10 am

parabolic problem-any help appreciated

October 25th, 2006, 6:40 pm

Hey guys, i`ve started in to a Masters and im having a few problems with numerical Algebra. I understand the fundamentals but we ve been giving a project which requires us to use matlab- and there my main problem arises. Ive never used matlab till this course, only maple and alot of it is going over my head. Ill list the question here and if any of you guys could give any help or direction i really would appreciate it!!!Consider the following class of singularly perturbed problemsAny help with particular reference to the matlab code would be like gold!! Thanks in advance I was asked to examine the numerical performance of different numerical methods applied to these problems, namely Explicit, implicit and centred differencing. I have discuss the stability of these schemes and make a plot of unstable output and pf stable output.
Last edited by bono06 on October 24th, 2006, 10:00 pm, edited 1 time in total.
 
User avatar
sheppardr
Posts: 2
Joined: March 22nd, 2005, 7:02 am

parabolic problem-any help appreciated

October 26th, 2006, 5:41 am

The fitting method in this paper by Daniel Duffy might be helpful.
Attachments
CrankNicolsonDuffy.zip
(1.9 MiB) Downloaded 80 times
 
User avatar
Cuchulainn
Posts: 23029
Joined: July 16th, 2004, 7:38 am

parabolic problem-any help appreciated

October 26th, 2006, 3:10 pm

QuoteConsider the following class of singularly perturbed problemsAny help with particular reference to the matlab code would be like gold!! Thanks in advance I was asked to examine the numerical performance of different numerical methods applied to these problems, namely Explicit, implicit and centred differencing. I have discuss the stability of these schemes and make a plot of unstable output and pf stable output.BonoThis article might be useful as well: FittingFDMThe coefficient of convection is 0 at x = 1/2 and it deserves special attention.You should google for researchers on SP problems: Miller, Farrell, O'Riordan, Hemker and in particular Shishkin (and his meshes!) These SP have been done by them.Traditional FDM is not robust for Sp problems P.S. are you in Dublin by any chance?
Last edited by Cuchulainn on October 25th, 2006, 10:00 pm, edited 1 time in total.
 
User avatar
bono06
Topic Author
Posts: 0
Joined: April 20th, 2006, 11:10 am

parabolic problem-any help appreciated

October 27th, 2006, 3:24 pm

Thanks guys for the help!!!The 2 articles are a good help, Yeah Cuchulainn im in dublin, Prof. O` riordan`s course; genuinely very interesting but my programming skills are less than apt hence my `appeal for help`
 
User avatar
Cuchulainn
Posts: 23029
Joined: July 16th, 2004, 7:38 am

parabolic problem-any help appreciated

October 27th, 2006, 4:10 pm

QuoteOriginally posted by: bono06Thanks guys for the help!!!The 2 articles are a good help, Yeah Cuchulainn im in dublin, Prof. O` riordan`s course; genuinely very interesting but my programming skills are less than apt hence my `appeal for help`Give my regards to all Edit: Do they organise Matlab programming courses for you there at the uni? It is only logical. Learning it on your own just takes too long.
Last edited by Cuchulainn on October 27th, 2006, 10:00 pm, edited 1 time in total.