boundary forcing: "LHS is constant but RHS is nonconstant along L axis."

[Jeffrey S. Oishi]

Hi all,

I'm having a bit of trouble doing boundary forcing with a GeneralFunction. I do the usual things: create a function that takes the generic *args and calls a function (bump, in this case, which just is a function in C^\inf_0(R)). I create a GeneralFunction on that, and make it parseable. Finally, I call it on the Boundary, but of course, this gives 

dedalus.tools.exceptions.SymbolicParsingError: LHS is constant but RHS is nonconstant along L axis.

It's fairly obvious why that's happening: the GeneralFuction is returning a 3D data cube, but of course a boundary condition should be a 2D array at the boundary of the third dimension. How can I work around this? I've seen some references to boundary forcing on this list, so I'm sure it's possible. 

Sample code (not all of it) here:

# Boundary condtion forcing

def BoundaryFunc(*args):

    V = args[0].data

    K = args[1].data

    hw = args[2].value

    xmax = args[3].value

    ampl = args[4].value

    

    return ampl*bump(V,hw,xmax)*bump(K,hw,xmax)

def B(*args, domain=domain,F=BoundaryFunc):

    return de.operators.GeneralFunction(domain,layout='g',func=F,args=args)

de.operators.parseables['BF'] = B

[stuff]

FP.add_bc("left(f) = 0")

FP.add_bc("right(f) = BF(V,K,hw,xmax,ampl)")

Any help would be most appreciated! After I get this working, I'll generate a simple example of boundary forcing for the docs. 

thanks,

Jeff

[Daniel Lecoanet]

Could you do

FP.add_bc("right(f) = right(BF(V,K,hw,xmax,ampl))")

?

--
You received this message because you are subscribed to the Google Groups "Dedalus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dedalus-users+unsubscribe@googlegroups.com.
To post to this group, send email to dedalu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/CAAOZTN_X%2BN7AMbVnm5cMkEg-VRBzSvmPZGYyF3PMATJTF1%2BoSg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

[Jeffrey S. Oishi]

Hi Daniel,

Oh, man. THAT'S embarrassing. Of course you can.

Thanks so much!

j

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

To post to this group, send email to dedalu...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/CAAOZTN_X%2BN7AMbVnm5cMkEg-VRBzSvmPZGYyF3PMATJTF1%2BoSg%40mail.gmail.com.
For more options, visit https://groups.google.com/d/optout.

--

You received this message because you are subscribed to the Google Groups "Dedalus Users" group.

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

To post to this group, send email to dedalu...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/CAJoYf%3Di%2BoUDohdR7c67QSoi%3DLmeT4p%3DNOnOA_maFVtEmj8SRfg%40mail.gmail.com.

[Eric Hester]

Hey Jeff,

Are some of your GeneralFunction arguments scalars? Because using scalar parameters as arguments I've just gotten errors saying " 'Scalar' object has no attribute 'layout' ".

I've also run into a similar error using the domain bases as arguments, since 'Array' objects don't have the layout attribute.

Is the simplest solution to just create fields for every GeneralFunction argument?

Cheers,

Eric

[Jeffrey S. Oishi]

Hi Eric,

Yes, the ones that use .value are scalars:

hw = args[2].value

I've also used the grids as arguments. I've attached a boundary forced wave equation script that has periodic in time and space forcing on the right boundary. It makes neat patterns; I haven't really thought much about whether or not it is correct, though. Nevertheless, it demonstrates how boundary forcing works. 

All, I'd like to propose this as a new example problem for dedalus. I'll issue a PR shortly, but if anyone wants to have a look, please let me know!

Let me know if this is helpful!

Jeff

--

You received this message because you are subscribed to the Google Groups "Dedalus Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dedalus-user...@googlegroups.com.
To post to this group, send email to dedalu...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/dedalus-users/26a06e6b-90fd-4725-9015-d88fbd3dfd98%40googlegroups.com.

[Eric Hester]

Thanks a bunch Jeff!

Much appreciated. I think something like this would be great for an example script.

Cheers,
Eric
> On 22 Sep 2017, at 10:16, Jeffrey S. Oishi <jso...@gmail.com> wrote:
>
> <boundary_forced_waves.py>