ffc team mailing list archive

Thread
Date

Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.

To: Mehdi Nikbakht <m.nikbakht@xxxxxxxxxx>
From: Kristian Oelgaard <k.b.oelgaard@xxxxxxxxx>
Date: Thu, 8 Jul 2010 09:42:13 +0100
Cc: FFC Mailing List <ffc@xxxxxxxxxxxxxxxxxxx>
In-reply-to: <1278576135.22552.9.camel@mech054.citg.tudelft.nl>

On 8 July 2010 09:02, Mehdi Nikbakht <m.nikbakht@xxxxxxxxxx> wrote:
> On Thu, 2010-07-08 at 08:50 +0100, Garth N. Wells wrote:
>>
>> On 08/07/10 08:33, Mehdi Nikbakht wrote:
>> > On Thu, 2010-07-08 at 08:22 +0100, Garth N. Wells wrote:
>> >> On Jul 8 2010, Anders Logg wrote:
>> >>
>> >>> On Wed, Jul 07, 2010 at 10:34:39PM +0100, Kristian Oelgaard wrote:
>> >>>> On 7 July 2010 20:22, Garth N. Wells<gnw20@xxxxxxxxx>  wrote:
>> >>>>>
>> >>>>>
>> >>>>> On 07/07/10 20:14, Anders Logg wrote:
>> >>>>>>
>> >>>>>> On Wed, Jul 07, 2010 at 06:26:20PM +0100, Kristian Oelgaard wrote:
>> >>>>>>>
>> >>>>>>> Supporting CellVolume makes it possible to do:
>> >>>>>>>
>> >>>>>>> CG = FiniteElement("Lagrange", triangle, 2)
>> >>>>>>> DG = FiniteElement("DG", triangle, 0)
>> >>>>>>> v = TestFunction(DG)
>> >>>>>>> f = Coefficient(CG)
>> >>>>>>> vol = triangle.v
>> >>>>>>
>> >>>>>> Would it be better to call it vol or volume instead of v? Or does it
>> >>>>>> have to be a one-letter word?
>> >>>>>>
>> >>>>>
>> >>>>> . . . or call it 'volume'.
>> >>>>
>> >>>> It can be whatever we want, I just followed what was already there.
>> >>>> Should we then rename 'd', 'n' and 'x' to 'geometric_dimension',
>> >>>> 'facet_normal', and 'spatial_coordinate' while we're at it?
>> >>>
>> >>> I think d, n, x are fine, but v does not necessarily look like a
>> >>> volume to me (it looks like a test function).
>> >>>
>> >>
>> >> I agree - d, n and x are all commonly used, but v for volume isn't.
>> >>
>> >> Garth
>> >
>> > I have another request related to geometric computation inside UFL. Is
>> > it possible to add a class called SurfaceNormal to geometry.py?
>> >
>>
>> I don't see how this is possible - it  can't be computed from the cell
>> geometry. Shouldn't it be a Constant/Coefficient computed by the user?
>
> Yes, it can. However, we need to define an extra finite element space
> for it. Now we have,
>
>  elem = VectorElement("DG", "triangle", 0)
>  n = Coefficient(elem)                       # Surface normal
>
> I would expect to have just one line for this,
>
>  n = SurfaceNormal("triangle")
>
> SurfaceNormal is just a coefficient defined on DG space with zero order.
> The same can be done for surface tangent.

This is something completely different, the FacetNormal and CellVolume
are geometric quantities which can be computed from cell information
alone (vertex coordinates) it does not rely on coefficients from any
finite element space. To compute the surface normal and surface
tangent, you would need extra information about the surface which is
not available in tabulate_tensor.

Kristian

> Mehdi
>
>
>
>>
>> Garth
>>
>> > Mehdi
>> >
>> >>
>> >>> --
>> >>> Anders
>> >>>
>> >>>
>> >>>> Kristian
>> >>>>
>> >>>>> Garth
>> >>>>>
>> >>>>>>
>> >>>>>>
>> >>>>>>> L = 1.0/vol*v*f*dx
>> >>>>>>>
>> >>>>>>> instead of using a Coefficient for 'vol' and then defining an
>> >>>>>>> Expression in DOLFIN.
>> >>>>>>>
>> >>>>>>> Currently, it will use the absolute value of the determinant of the
>> >>>>>>> Jacobian in tabulate_tensor ( std::abs(detJ) ).
>> >>>>>>>
>> >>>>>>> Kristian
>> >>>>>>>
>> >>>>>>> On 7 July 2010 18:19,<noreply@xxxxxxxxxxxxx>   wrote:
>> >>>>>>>>
>> >>>>>>>> ------------------------------------------------------------
>> >>>>>>>> revno: 1498
>> >>>>>>>> committer: Kristian B. Ølgaard<k.b.oelgaard@xxxxxxxxx>
>> >>>>>>>> branch nick: ffc
>> >>>>>>>> timestamp: Wed 2010-07-07 18:17:07 +0100
>> >>>>>>>> message:
>> >>>>>>>>   Added support for CellVolume from UFL.
>> >>>>>>>> modified:
>> >>>>>>>>   ChangeLog
>> >>>>>>>>   ffc/quadrature/optimisedquadraturetransformer.py
>> >>>>>>>>   ffc/quadrature/quadraturetransformer.py
>> >>>>>>>>   ffc/quadrature/quadraturetransformerbase.py
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>> Your team FFC Core Team is subscribed to branch lp:ffc.
>> >>>>>>>> To unsubscribe from this branch go to
>> >>>>>>>> https://code.launchpad.net/~ffc-core/ffc/main/+edit-subscription
>> >>>>>>>>
>> >>>>>>>> === modified file 'ChangeLog' --- ChangeLog   2010-07-01 17:43:36
>> >>>>>>>> +0000 +++ ChangeLog   2010-07-07 17:17:07 +0000 @@ -1,3 +1,4 @@ + -
>> >>>>>>>> Added support for new geometric quantity CellVolume in UFL.  0.9.3
>> >>>>>>>> [2010-07-01]  - Make global_dimension for Real return an int
>> >>>>>>>> instead of double, bug # 592088  - Add support for facet normal in
>> >>>>>>>> 1D.
>> >>>>>>>>
>> >>>>>>>> === modified file 'ffc/quadrature/optimisedquadraturetransformer.py'
>> >>>>>>>> --- ffc/quadrature/optimisedquadraturetransformer.py    2010-07-01
>> >>>>>>>> 17:41:40 +0000
>> >>>>>>>> +++ ffc/quadrature/optimisedquadraturetransformer.py    2010-07-07
>> >>>>>>>> 17:17:07 +0000
>> >>>>>>>> @@ -187,7 +187,7 @@
>> >>>>>>>>          return {():new_val}
>> >>>>>>>>
>> >>>>>>>>      #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>> -    # FacetNormal (geometry.py). +    # FacetNormal, CellVolume
>> >>>>>>>> (geometry.py).     #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>>      def facet_normal(self, o,  *operands):         #print("Visiting
>> >>>>>>>> FacetNormal:") @@ -216,6 +216,17 @@
>> >>>>>>>>
>> >>>>>>>>          return {(): create_symbol(normal_component, GEO)}
>> >>>>>>>>
>> >>>>>>>> +    def cell_volume(self, o,  *operands): +        # Safety
>> >>>>>>>> check. +        ffc_assert(not operands, "Didn't expect any
>> >>>>>>>> operands for FacetNormal: " + repr(operands)) + +        # FIXME:
>> >>>>>>>> KBO: This has to change for higher order elements +        detJ =
>> >>>>>>>> format["det(J)"](self.restriction) +        volume =
>> >>>>>>>> format["absolute value"](detJ) +        self.trans_set.add(detJ) +
>> >>>>>>>> +        return {():create_symbol(volume, GEO)} +     def
>> >>>>>>>> create_argument(self, ufl_argument, derivatives, component,
>> >>>>>>>> local_comp,                   local_offset, ffc_element,
>> >>>>>>>> transformation, multiindices):         "Create code for basis
>> >>>>>>>> functions, and update relevant tables of used basis."
>> >>>>>>>>
>> >>>>>>>> === modified file 'ffc/quadrature/quadraturetransformer.py'
>> >>>>>>>> --- ffc/quadrature/quadraturetransformer.py     2010-06-30 09:59:49
>> >>>>>>>> +0000
>> >>>>>>>> +++ ffc/quadrature/quadraturetransformer.py     2010-07-07 17:17:07
>> >>>>>>>> +0000
>> >>>>>>>> @@ -262,7 +262,7 @@
>> >>>>>>>>          return {():f_abs(operands[0][()])}
>> >>>>>>>>
>> >>>>>>>>      #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>> -    # FacetNormal (geometry.py). +    # FacetNormal, CellVolume
>> >>>>>>>> (geometry.py).     #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>>      def facet_normal(self, o,  *operands):         #print("Visiting
>> >>>>>>>> FacetNormal:") @@ -291,6 +291,18 @@
>> >>>>>>>>
>> >>>>>>>>          return {():normal_component}
>> >>>>>>>>
>> >>>>>>>> +    def cell_volume(self, o,  *operands): +        # Safety
>> >>>>>>>> check. +        ffc_assert(not operands, "Didn't expect any
>> >>>>>>>> operands for CellVolume: " + repr(operands)) + +        # FIXME:
>> >>>>>>>> KBO: This has to change for higher order elements +        detJ =
>> >>>>>>>> format["det(J)"](self.restriction) +        volume =
>> >>>>>>>> format["absolute value"](detJ) +        self.trans_set.add(detJ) +
>> >>>>>>>> +        return {():volume} + +     def create_argument(self,
>> >>>>>>>> ufl_argument, derivatives, component, local_comp,
>> >>>>>>>> local_offset, ffc_element, transformation, multiindices):
>> >>>>>>>> "Create code for basis functions, and update relevant tables of
>> >>>>>>>> used basis."
>> >>>>>>>>
>> >>>>>>>> === modified file 'ffc/quadrature/quadraturetransformerbase.py'
>> >>>>>>>> --- ffc/quadrature/quadraturetransformerbase.py 2010-07-01 17:41:40
>> >>>>>>>> +0000 +++ ffc/quadrature/quadraturetransformerbase.py 2010-07-07
>> >>>>>>>> 17:17:07 +0000 @@ -250,12 +250,16 @@         error("This object
>> >>>>>>>> should be implemented by the child class.")
>> >>>>>>>>
>> >>>>>>>>      #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>> -    # FacetNormal (geometry.py). +    # FacetNormal, CellVolume
>> >>>>>>>> (geometry.py).     #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>>      def facet_normal(self, o,  *operands):         print
>> >>>>>>>> "\n\nVisiting FacetNormal: ", repr(o)         error("This object
>> >>>>>>>> should be implemented by the child class.")
>> >>>>>>>>
>> >>>>>>>> +    def cell_volume(self, o,  *operands): +        print
>> >>>>>>>> "\n\nVisiting CellVolume: ", repr(o) +        error("This object
>> >>>>>>>> should be implemented by the child class.") +     #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>>      # Things that can be handled by the base class.     #
>> >>>>>>>> -------------------------------------------------------------------------
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>>
>> >>>>>>>
>> >>>>>>> _______________________________________________
>> >>>>>>> Mailing list: https://launchpad.net/~ffc
>> >>>>>>> Post to     : ffc@xxxxxxxxxxxxxxxxxxx
>> >>>>>>> Unsubscribe : https://launchpad.net/~ffc
>> >>>>>>> More help   : https://help.launchpad.net/ListHelp
>> >>>>>>>
>> >>>>>>>
>> >>>>>>> _______________________________________________
>> >>>>>>> Mailing list: https://launchpad.net/~ffc
>> >>>>>>> Post to     : ffc@xxxxxxxxxxxxxxxxxxx
>> >>>>>>> Unsubscribe : https://launchpad.net/~ffc
>> >>>>>>> More help   : https://help.launchpad.net/ListHelp
>> >>>>>
>> >>>
>> >>
>> >> _______________________________________________
>> >> Mailing list: https://launchpad.net/~ffc
>> >> Post to     : ffc@xxxxxxxxxxxxxxxxxxx
>> >> Unsubscribe : https://launchpad.net/~ffc
>> >> More help   : https://help.launchpad.net/ListHelp
>> >
>
>
> _______________________________________________
> Mailing list: https://launchpad.net/~ffc
> Post to     : ffc@xxxxxxxxxxxxxxxxxxx
> Unsubscribe : https://launchpad.net/~ffc
> More help   : https://help.launchpad.net/ListHelp
>

References

Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Kristian Oelgaard, 2010-07-07
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Anders Logg, 2010-07-07
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Garth N. Wells, 2010-07-07
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Kristian Oelgaard, 2010-07-07
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Anders Logg, 2010-07-08
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Garth N. Wells, 2010-07-08
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Mehdi Nikbakht, 2010-07-08
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Garth N. Wells, 2010-07-08
Re: [Branch ~ffc-core/ffc/main] Rev 1498: Added support for CellVolume from UFL.
From: Mehdi Nikbakht, 2010-07-08