Strange chebmatrix BVP behavior

[Christopher Marcotte]

Hi everyone!

I've run into a strange issue with solving BVPs using the chebmatrix syntax. Rather, the BVP is solved fine (see demonstrative script below) but the solver fails when checking the boundary residuals. The error is "Too many input arguments."

Is this a known issue? I could't find anything quite matching the description in the group or in the open issues on Github.

The reason I'd like to cast the problem as a chebmatrix is because eventually I'd like to use the followpath functions to do some simple continuation, alternatively if you know how to use this function with the first form, I'd love to see it.

CM

[Nick Hale]

I think the offending line needs to be replaced with 

        if ( cellArg )

            bcU = N.bc(x, u);

        else

            bcU = N.bc(x, uBlocks{:});

        end

(see the similar sections for N.rbc and N.lbc.)

I've issued a pull request:

https://github.com/chebfun/chebfun/pull/2057

N

--
You received this message because you are subscribed to the Google Groups "chebfun-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to chebfun-users+unsubscribe@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/chebfun-users/8d4b8b63-0bf9-4ca4-ae39-6f65210c487a%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Christopher Marcotte]

Thank you, Nick!

On Friday, February 10, 2017 at 3:10:37 PM UTC, Nick Hale wrote:

I think the offending line needs to be replaced with 

        if ( cellArg )

            bcU = N.bc(x, u);

        else

            bcU = N.bc(x, uBlocks{:});

        end

(see the similar sections for N.rbc and N.lbc.)

I've issued a pull request:

https://github.com/chebfun/chebfun/pull/2057

N

On 10 February 2017 at 16:44, Christopher Marcotte <christophe...@gmail.com> wrote:

Hi everyone!

I've run into a strange issue with solving BVPs using the chebmatrix syntax. Rather, the BVP is solved fine (see demonstrative script below) but the solver fails when checking the boundary residuals. The error is "Too many input arguments."

Is this a known issue? I could't find anything quite matching the description in the group or in the open issues on Github.

The reason I'd like to cast the problem as a chebmatrix is because eventually I'd like to use the followpath functions to do some simple continuation, alternatively if you know how to use this function with the first form, I'd love to see it.

CM

--
You received this message because you are subscribed to the Google Groups "chebfun-users" group.

To unsubscribe from this group and stop receiving emails from it, send an email to chebfun-user...@googlegroups.com.

[Ron Shvartsman]

Hi everyone,

I'm having the same issue, but the proposed fix hasn't helped. I've attached below my super short script, which attempts to perform continuation on a chebmatrix, but I get the same issue as Christopher when declaring my boundary conditions. I'd appreciate it if someone could take a look!!

Ron

[Nick Hale]

Hi Ron

I've had a look, but I do not think there is a quick fix here. followpath() 

does not seem to currently support systems of ODEs.

(For anyone wishing to investigate further, it seems that a) the way 

followpath concatenates the dependent variable with the bifurcation

parameter breaks a lot of things in linearize and feval, b) much of

followpath and tangentBVP seem to be written with only scalar

problems in mind.)

 

Sorry that I can't be more of more help.

Nick

[Ron Shvartsman]

Thank you for the response Nick!