320. 310. 304.9998]
> shouldn't the first and last value be (near) 400. and 300., respectively?
>
Depends what the boundary conditions are. We assume that the gradient is
zero on the edge. This transfers to the natural boundary condition when
solving se
-- Forwarded message --
From: Daniel Wheeler
Date: Fri, Dec 2, 2011 at 2:39 PM
Subject: Re: 3D Level Set Method
To: David Ely
Hi David,
No 3D level set and no easy way to make it happen.
Truth be told, the level set implementation doesn't really belong as part
of fip
On Wed, Nov 9, 2011 at 10:10 AM, Daniel Wheeler
wrote:
> On Thu, Nov 3, 2011 at 11:47 AM, Douwe Bruinsma wrote:
>
>> I'm impressed by the FiPy package, but don't understand how to implement
>> the convection boundary conditions for my problem.
>>
>
> Hi
...and again.
-- Forwarded message --
From: Daniel Wheeler
Date: Tue, Nov 1, 2011 at 11:12 AM
Subject: Re: open virtually the graphics interface
To: Julien Derr
Julien,
Try the script below and see if it works for you. Also comment out this
line <
http://matforge.
Messed up the reply again.
-- Forwarded message --
From: Daniel Wheeler
Date: Tue, Nov 1, 2011 at 11:18 AM
Subject: Re: Simple 2D Problem
To: Lorenzo Isella
I think Gmsh should be able to do this. Try it. Tale a look at <
http://matforge.org/fipy/browser/trunk/examp
1
> return self
>
>
> Interface_Variable = CellVariable(name="InterfacecellvvariableI",
> mesh=mesh,value=0.,hasOld=1)
> Interface_Variable=classeinterface(phi)
An __init__ is needed that calls CellVariable's __init__ and specifies
the phi dependence. See
variable.
Another option is to create a cell variable class with a calcValue
method that uses _cellInterfaceFlag to update its own value and also
requires phi. It will then update whenever phi changes.
--
Daniel Wheeler
___
fipy mailing list
fipy@nist.gov
dle.
> define the BC in the "infinity" using getExteriorFaces()
> the problem: how can i define the robin BC on the inner hole-sphere
> surface?
Identify the interior faces using (exteriorFaces & r < R) and make a
boundary condition.
> or is there a better way?
the
> outside boundary of the mesh.
Julien, As far as I can tell, you were printing cell indices at face
locations so what you are seeing is gobbledygook. Have you checked the
values using the correct indices?
--
Daniel Wheeler
___
fipy mailing list
f
nt of how to
implement this idea with the second mesh. I'm not sure I got it. I
also have to admit that I'm totally new to FiPy.
Thanks in advance,
Samo
p.s. Next time I'll make sure to post to FiPy mailing list.
On Oct 4, 2011, at 4:59 PM, Daniel Wheeler wrote:
> Hi Samo, Th
uld make a CellVariable non-contiguous, since then it wouldn't
be a CellVariable.
On Wed, Sep 21, 2011 at 4:02 PM, list user wrote:
> When constructing a CellVariable from a lxnXm mesh am I guaranteed the data
> to be contiguous?
>
--
Daniel Wheeler
x27;t seem to find a method to set
> a _pulse_ of concentration.
You can just index as you would with a numpy array.
Hope that helps.
Cheers.
--
Daniel Wheeler
!! is there a reason why the end of the mesh would
> have the same property as the interface?
Don't know off hand.
> does the self.distanceVar._cellInterfaceFlag Flag is also one at the
> outside edge of the mesh ??
Print the values and check it.
--
Daniel Wheeler
t; for explicit method); and in that case, where is my mistake ?
You have the right idea. Make sure you initialize the time variables
with a float rather than an integer, but I don't think that should be
a problem
> - or is it something else, and in that case, could someone please put me on
> the right track?
Post the entire script so that it runs and I'll try and figure it out.
--
Daniel Wheeler
ent term coefficient should just be a cell value. So simply
use "r**2" for the transient term.
--
Daniel Wheeler
=0? It's a regular 1D mesh, not cylindrical.
--
Daniel Wheeler
## mesh.getCellCenters() for older fipy
>>> myRegularEqn + ImplicitSourceTerm(1e+20 * (x > 3)) - 1e-20 * (x
> 3) * value == 0
This would set the value of the variable to be `value` where x > 3.
--
Daniel Wheeler
bla c \right) + S_P c = 0
where $S_P$ is a large value.
--
Daniel Wheeler
DiffusionTermNoCorrection.
Mixing DiffusionTermNoCorrection and DiffusionTerm won't help, but I a
not sure that there is a problem with the solution anyway.
--
Daniel Wheeler
_p and S_c large and have the correct ratio in a selected
region. It's probably best to do this directly in the script by adding
to the existing equations rather than messing directly with the
internals of FiPy.
The above is a for a fixed value. An internal flux is a little different.
Hope this helps.
--
Daniel Wheeler
/diffusionTermNoCorrection.py>
and in the latest release, it's in"fipy.terms.diffusionTerm"
<http://matforge.org/fipy/browser/tags/version-2_1_2/fipy/terms/diffusionTerm.py>.
I believe you need to import this directly (not auto-imported with
"from fipy import *"). Hope this helps.
Cheers
--
Daniel Wheeler
it on
version-2_1 since we are working towards a new release.
> Please check attacked file.
None of those failures are important. Two are vtk, which you probably
don't have and don't need most likely and the others seem to be
testing failures that fail in different ways. Probably because your
version of numpy is quite new, but nothing to worry about.
Cheers
--
Daniel Wheeler
e.cacheMatrix()
e.justResidualVector(v)
print e.matrix
> I would like to formulate the problem in fipy (because later I will use it
> for time evolution) but need to solve for the static configurations first.
> (Also, is there a way of not using sparse solvers and just using numpy
> arrays? For debugging small systems this would be very useful...)
You can ask for "e.matrix.numpyArray" to get the matrix as a numpy
array if you prefer that format.
Hope this helps. Cheers.
--
Daniel Wheeler
"/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/FiPy-2.1-py2.7.egg/fipy/meshes/numMesh/mesh.py",
> line 521, in _calcFaceCellToCellNormals
> orientation = 1 - 2 * (numerix.dot(self.faceNormals,
> self.faceCellToCellNormals) < 0)
> File
> "/opt/local/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/site-packages/FiPy-2.1-py2.7.egg/fipy/tools/numerix.py",
> line 835, in dot
> return a1.dot(a2)
> ValueError: matrices are not aligned
>>>>
>
>
--
Daniel Wheeler
x". For this you could use
justResidualVector. However, this really defeats the purpose of using
FiPy as you will no longer be using an implicit scheme, but an
explicit scheme. You also might not want the transient term in the
FiPy equation definition since odeint deals with this. Not sure how to
get around the implicit/explicit issue though.
--
Daniel Wheeler
On Mon, Jun 20, 2011 at 11:11 AM, Daniel Wheeler
wrote:
> Hi Rajeev,
>
> People have asked about this on the list have a given indications that
> they are solving the necessary equations sucessfully. However, we
> don;t have a ny canned examples for elasticity to get you strat
pe memory alloys. As such, I need to solve the
> mechanical equilibrium and elasticity equations. I was wondering if the
> elasticity equation has been solved before using FiPy.
> Thank you!
> Sincerely,
> Rajeev
--
Daniel Wheeler
t; 0.2)
integral = (mask * N * mesh.getCellVolumes()).sum()
for example.
> I would
> also like the result be of ndarray type (i.e., representing the
> integrated value at each xc).
Not sure what you are asking. The integral is obviously independent of location.
Hope the above helps.
--
Daniel Wheeler
tp://matforge.org/fipy/browser/trunk/fipy/meshes/gmshImport.py#L297>
Hence info associated with faces will be lost. Sorry I can't be of more help.
--
Daniel Wheeler
tput is:
>
> [ 0.5 1.5 2.5 3.5 4.5 5.5 6.5 7.5 8.5 9.5]
>
> Could it be that a factor of pi is missing here? Assuming that the half
> cylinder
> is modelled...
Hi Burak, The assumption is a 1 Rad wedge. For the inner working of
FiPy the factor is not important. Hope this helps.
--
Daniel Wheeler
d to the data-structures used in FiPy?
> Best,
> Biswa
--
Daniel Wheeler
ght be worth trying different solvers. Try using the LinearLUSolver
for the equations that are throwing the iteration warning or switching
to trilinos's gmres solver as an alternative.
--
Daniel Wheeler
hed the code, if someone get the courage and the time to look at it
> ...or do you have ideas just like that ?
Julien, I tried running the code, but I didn't get any of those
warnings with trunk or version-2_1. Version-2_1 failed with an
attribute error. Sorry.
--
Daniel Wheeler
.
>
> Kind regards,
>
> Christine
>
>
--
Daniel Wheeler
t; v = CellVariable(mesh=m)
>>> r = (2.9, 2.9)
>>> v[argmin(sum((r - m.cellCenters)**2))] = 100.
>>> print v
[ 0.0.0.0.0.0.0.0. 100.]
> My question is: is there a function that gives me the index for a given mesh
> and position r?
No, but argmin allows you to do this fairly easily.
--
Daniel Wheeler
wo weeks ago FiPy on the web and I'm trying to use it
> for my purposes. I'm currently trying some simple things to understand a
> little bit
> more how using FiPy. My conclusion is that Fipy is very easy to use and
> seems
> powerful.
> You really did a great job !
>
ot float64 in both cases !!
> May my *.py file help you ?
>
> Thanks
>
> Marc
>
> Daniel Wheeler a écrit :
>
>>
>> What do you get if you print numerix.array(S).dtype?
>>
>> On Thu, May 19, 2011 at 9:13 AM, Marc Saudreau
>> wrote:
>>>
> Thanks again for your help and time
>
> Marc
>
>
> Daniel Wheeler a écrit :
>
>>
>> Hi Marc,
>>
>> CellVariable's __call__ method extracts data. So you can do the following:
>>
>> from fipy import *
>> import pylab
>>
>> L
; Hi everyone,
>
> I'm trying to use Fipy to solve a 2D heat transfer problem, and I need to
> extract from my 2D grid, some 1D profiles. For instance I need to plot
> temperature versus y axis for x given.
> I had a look at Fipy examples but I did not find any way to do it.
> Thanks a lot for any help.
>
> Best regards,
>
> Marc
>
>
--
Daniel Wheeler
time() <= 20:
> for step in range(100):
> time.setValue(time() + dt)
> # for a new boundary condition for source term (#2)
> grad2setValue(BC_Left, where=mesh.getFacesLeft())
> grad2.setValue(BC_Right, where=mesh.getFacesRight())
> # solve the equations
> eqX.
On Wed, Apr 27, 2011 at 3:50 AM, Gyeong-Geun LEE wrote:
> Thanks for your suggestion.
> However, it did not work.
> The values at boundary were diverged!
Post your script and I'll take a look.
--
Daniel Wheeler
p in range(100):
> time.setValue(time() + dt)
> eqX.solve(var=phi, boundaryConditions=BCs, dt=dt)
> eqX2.solve(var=phi2, dt=dt)
> # for a new boundary condition
> grad.setValue(BC_Left, where=mesh.getFacesLeft())
> grad.setValue(BC_Right, where=mesh.getFacesRight())
> if __name__ == '__main__':
> viewer.plot()
>
> # Final check
> print phi[0:3]
> print phi2[0:3]
> raw_input('Calculation done. Please hit any key...')
>
>
>
--
Daniel Wheeler
to type
$ gmsh
at the command line and have it launch. That should be enough. The
examples you were running all use gmsh for creating their meshes.
--
Daniel Wheeler
On Fri, Apr 1, 2011 at 3:56 AM, Jacek Hoffman wrote:
> Package python-mpi4py ver. 1.2.2-1 is sent to Daniel
Thanks! I'll see if it helps. I'll take it down if mpi4py ever gets an
up to date deb.
--
Daniel Wheeler
hanks again very much for your very valuable help!!!
I'm glad you are making some progress.
--
Daniel Wheeler
On Thu, Mar 31, 2011 at 1:47 PM, Jacek Hoffman wrote:
> 4 -- installing package system-wide as root (in such a case there is no
> need to worry about PYTHONPATH or LD_LIBRARY_PATH, I fancy):
> dpkg -i python-mpi4py_1.2.2-1~pre1_i386.deb
Shall we put this on the download page?
-
pi4py" since once there wasn't
> "easy_uninstall".
> I prefer using package manager so I did it the hard way.
Appreciate the effort and thanks for posting the recipe.
--
Daniel Wheeler
had to build it manually. I
haven't actually tried for awhile, but I'm planning to leave well
alone until we update to sqeeze.
Hope this helps in some way.
--
Daniel Wheeler
cript.py --pypsparse")? Ideally, that would never even attempt any
Trilinos or mpi4py imports, but I think it does. I think we should
change it so it doesn't. It is often the case that Trilinos breaks in
some way so even importing in a try-except block breaks things.
Cheers
--
Daniel Wheeler
Set/electroChem/metalIonSourceVariable.py#L88>
with "return self.distanceVar._cellInterfaceFlag * 1e+20". You
probably want to construct the metal equation in your script to avoid
editing these files directly. That will make c=0 on the interface
(phi=0).
> v= \nabla c
I think we've already discussed the above condition.
--
Daniel Wheeler
//www.ctcms.nist.gov/fipy/fipy/generated/models.levelSet.electroChem.html#module-fipy.models.levelSet.electroChem.metalIonDiffusionEquation>?
In the examples in FiPy there are essentially two conditions at the
interface. It seems like both are required. What is going on
physically? Is the metal being deposited at the interface? Do you need
to conserve mass at the interface?
--
Daniel Wheeler
ing grid type meshes in
Gmsh. Removing an interior circular region is certainly possible in
Gmsh. I think you just need to define an interior circular region with
four circular lines and then define the line loops with the correct
ordering and direction. See the Gmsh tutorial for this
<http://
s is the
"self.distanceVar.cellInterfaceAreas" field, which gives an
approximation for the area of the boundary interface in a given cell.
Hope this helps.
--
Daniel Wheeler
amples at
>> >> http://matplotlib.sourceforge.net/ ?
>> >>
>> >> You might need to experiment with different matplotlib backends:
>> >>
>> >> http://matplotlib.sourceforge.net/faq/installing_faq.html#backends
>>
>>
>>
>>
>
>
--
Daniel Wheeler
ate, not sure why. Also the
example is really set up for deposition rather than dissolution so I
added the abs().
--
Daniel Wheeler
l run it on my machine?
> and also my installation of spyparse looks weird because still a problem a
> running test_spmatrix.py
If the fipy tests pass then don't worry about it.
--
Daniel Wheeler
a problem, just set a very large source wherever D=0.
--
Daniel Wheeler
can explain
that.
> case of growth, which means if my equation is in the following form:
> \frac{\partial b}{\partial t} = D \frac{\partila ^2 b}{\partial x^2} +
> rb(1-b/k)
> where r and k are constants.
That is a different equation from above. Do you have a problem solving
it or posing it in FiPy?
--
Daniel Wheeler
ry so that
you can run
$ python examples/levelSet/electroChem/simpleTrenchSystem.py
then try
$ set PYTHONPATH=.
$ python -c 'import fipy; print fipy.__version__'
What do you get? Is it different from 2.2-dev?
$ python examples/levelSet/electroChem/simpleTrenchSystem.py
Does this now work?
--
Daniel Wheeler
Sorry.
On Fri, Mar 4, 2011 at 5:39 PM, Daniel Wheeler
wrote:
> You can check and see that the example you are trying to run don't not
> match your installed version of FIPy by importing the Grid2D instance
> and looking at the arguments it takes.
You can check and see that the e
we express that
> in terms of boundary conditions?
Sorry. Can't really help with that. We need equations and boundary
conditions to make progress.
--
Daniel Wheeler
u are running appear to be more recent than
the version of fipy you are using.
--
Daniel Wheeler
;
>>> On Feb 24, 2011, at 8:20 AM, Julien Derr wrote:
>>>
>>> > which installation of pysparse would you suggest ?
>>> > I have currently 1.1.1
>>>
>>> I am successfully running '1.2-dev213' on a Mac OS X Snow Leopard system
>>> and '1.2-dev' on a Debian lenny system.
>>>
>>> I believe in both cases that I just did `easy_install pysparse`.
>>>
>>> How are you building it? What is your system?
>>>
>>>
>>>
>>
>
>
--
Daniel Wheeler
ionRateVariable is a scalar quantity. You may
want to advect the interface proportinal to the gradient of the metal
ion field. To do this you might set,
depositionRateVariable = metalVar.getGrad().dot(distanceVar.getGrad())
I think this is probably what you want.
> thanks a lot for your help,
No problem.
Cheers
--
Daniel Wheeler
gt; numberOfsteps=10
> cflNumber = 0.2
> for step in range(numberOfsteps):
> viewer.plot()
>
> phi.calcDistanceFunction()
>
> phi.updateOld()
> c.updateOld()
>
> phi.extendVariable(c.getFaceGrad())
>
> dt = cflNumber * cellSize / c.getFaceGrad().max()
>
> advectionEquation.solve(phi, dt=dt)
>
> ceq.solve(var=c, dt=dt,boundaryConditions=BCs)
>
>
>
>
>
--
Daniel Wheeler
hod
for installing <http://matforge.org/fipy/downloader/download/release/16>.
--
Daniel Wheeler
On Wed, Feb 16, 2011 at 9:08 AM, Daniel Wheeler
wrote:
> You're obviously using the matplotlib viewer and not the mayavi
> viewer. FiPy selects one or the other. You can specify this by setting
> the "FIPY_VIEWER" environment variable to "matplotlib" or "
>> import matplotlib
>> print matplolib.__version__
>>
>> ?
>>
>>
>> You can try editing line 124 of
>> C:\Python25\lib\site-packages\fipy\viewers\matplotlibViewer\matplotlib2DGridViewer.py
>> from
>>
>> self.image.clim(vmax=datamax, vmin=datamin)
>>
>> to
>>
>> self.image.set_clim(vmax=datamax, vmin=datamin)
>>
>>
>>
>
>
--
Daniel Wheeler
pecify n\dot\vec{J} with the new scheme.
> My guess is that phi.faceGrad.constrain(value=mesh._getFaceNormals() *
> phi_grad_mag, where=...) would do the trick, but Wheeler may know some
> finite-volumy reason for why that's not necessary.
> It looks to me like phi.faceGrad.constrain() has only been tested in 1D
> cases, where this doesn't matter. Again, Wheeler will be happy to look into
> this when he gets back... won't you Wheeler?
I'll change the examples to use vector values.
--
Daniel Wheeler
3, ny=3)
>>> x1, y1 = m1.getCellCenters()
>>> v1 = CellVariable(mesh=m1, value=x1 * y1)
>>> m0 = Grid2D(nx=2, ny=2)
>>> v0 = CellVariable(mesh=m0, value=v1(m0.getCellCenters()))
I haven't checked the above, so possible typos.
--
Daniel Wheeler
75 0.75 2.25]
> when I look at newphi, there are only zeros, it didn't take the values of
> oldphi; isn't it supposed to do that with setValue ?
What happens when you "print oldphi((xc,yc))"? Is it actually non zero
and is it the same length as phi? You should also be able to do
newphi.setValue(oldPhi) if they are the same length.
--
Daniel Wheeler
there a simple way to get phi(xa,ya) where xa and ya is any arbitrary
> value (ie not on a mesh ) ?
X and Y can have arbitrary values. They can also be arbitrary array values.
--
Daniel Wheeler
ze}; Circle(6) = {2, 1, 3};
> Circle(7) = {3, 1, 4};Circle(8) = {4, 1, 5}; Circle(9) = {5, 1, 2}; Line
> Loop(10) = {6, 7, 8, 9};Plane Surface(12) = {10}; \'\'\' % locals() "
>
> and then define
>
> mesh = GmshImporter2D(cmdbase)
>
>
> doesn't seem to work ...
>
> thanks for your help!
>
> Julien
>
>
--
Daniel Wheeler
r = 10
> phase[((x - cx)**2 + (y - cy)**2 + (z - cz)**2) < r**2]= 1
> for i in range(particles):
> cx = random.random()*nx*dx
> cy = random.random()*nx*dx
> cz = random.random()*nx*dx
> r = 10
> phase[((x - cx)**2 + (y - cy)**2 + (z - cz)**2) < r**2]= 0
> print "pore volume fraction=",phase.getCellVolumeAverage()
> -
>
>
--
Daniel Wheeler
gt; The program stops to respond and I am having to forcefully close the entire
> application.
>
> I am running fipy 2.1.1 along with Enthought python distribution 6.3.1 which
> comes with python 2.6 on a windows - 64bit machine.
>
> Any help will be greatly appreciated.
>
> Thanks
> Ram.
>
--
Daniel Wheeler
"equal to" symbol.
>
> Am I missing something here?
>
> Thanks
> Ram.
>
--
Daniel Wheeler
t;import enthought" and from "enthought import mayavi" works at the command
> line. The commands run without producing errors. But would that make fipy
> recognize mayavi when we run "python setup.py test" from the dos prompt?
>
> Ram.
>
> On Mon, Nov 15, 2
; been installed. All the modules that fipy identifies are found in
> C:\Python26\Lib\site-packages\. I know that EPD comes with mayavi and can be
> found in C:\Python26\Lib\site-packages\enthought\. How do I make fipy
> recognize mayavi which is in a different folder?
>
> Thanks
> Ram.
>
--
Daniel Wheeler
ase.finalize_options(self)
> File
> "/home/lucia/LuciaDocuments2010/Lucia2010/FiniteElementMethod/FiPy-2.1/setuptools-0.6c11-py2.6.egg/setuptools/command/test.py",
> line 64, in finalize_options
>
> AttributeError: Distribution instance has no attribute 'test_suite'
>
>
> Ok, that's all. FiPy is working fine for me now if I run it from its
> own directory. I am using version 2.1. I was going to try 2.1.1
> yesterday before sending mail but the site was broken for downloading
> :-)
>
> Thank you very much for your help.
> Lucia
>
>
--
Daniel Wheeler
spective code and the viewers themselves are installed
> (gnuplot, matplotlib and mayavi2).
>
> I don't know how to have FiPy recognize the viewers.
>
> Thanks a lot for your help,
> Lucia
>
>
--
Daniel Wheeler
e throw some light on it.
> Your help will be greatly appreciated.
See if FiPy works, probably the best way to know. Don't worry about
the "tar.gz" if you have already installed with the the ".exe".
Cheers
--
Daniel Wheeler
bit and Red hat linux platforms. I see the installation files for
> windows-32 bit and for Mac-OS but I don't see them for windows - 64bit and
> Redhat linux.
>
> Any help will be appreciated.
>
> Thanks
>
> Ram Balachandran
>
>
> Ram Balachandran, PhD
> Senior R&D Engineer
> Applied Technologies
> Boston Scientific
> Tel: 763 255 0208
--
Daniel Wheeler
ces"
>
> Is there a way to enforce these complicated boundaries in fipy?
>
You can't have interior boundary conditions in FiPy. Cast the boundary
condition to a volume integral and apply a source term based on location or
use Benny's suggestion, which can more accurately capture the boundary.
--
Daniel Wheeler
trilinos matrix, which can be improved
on, but that will take more time. I will try and merge these changes
back to version-2_1 ASAP and maybe do another release.
Cheers
On Thu, Oct 7, 2010 at 12:55 PM, Daniel Wheeler
wrote:
>
> Erik, FYI <http://matforge.org/fipy/browser/branches/e
ooks to me that only, a scalar, or a constant vector is allowed in this
> term. Any suggestion?
The coeff for ConvectionTerm needs to be a vector. Try,
>>> cv = CellVariable(mesh = m, rank=1)
>>> __ConvectionTerm(coeff = cv)
"rank=1" makes the variable a vector value.
--
Daniel Wheeler
ion term with a variable coefficient? If so, the answer
is yes. The coefficient can be any vector field that changes in space
and time, but the term can not be posed to FiPy in the form you gave.
Hope the above helps.
--
Daniel Wheeler
aptive"
> time-stepping loop that I implemented.
>
> Best,
>
> Erik
>
> On Oct 5, 2010, at 3:09 PM, Daniel Wheeler wrote:
>
>> Hi Erik, I've attached a script that seems to run faster. It took
>> 57.60s to do 200 steps as opposed to your original script, which t
initial residual, rather than some arbitrary number.
>>
>> On Sep 30, 2010, at 12:55 PM, Erik Sherwood wrote:
>>
>>>
>>> Thanks for taking a look at the code. I'm using FiPy 2.1.
>>>
>>> Erik
>>>
>>> On Sep 30, 2010, at 11
; print "Step %d"%(step)
[snip]
> # Update the plots
> if __name__ == '__main__':
> Uviewer.plot()
> Vviewer.plot()
> Wviewer.plot()
--
Daniel Wheeler
while residual > desiredResidual and sweepcount < 300:
> #print "Sweep %d"%(sweepcount+1)
> residual = eqn.sweep(var=var, dt=dt)
> sweepcount += 1
> # Bail if residual remains too large
> if residual > desiredResidual:
> print "Unable to reduce residual!"
> break
> if residual > desiredResidual:
> break
>
> # Update the plots
> if __name__ == '__main__':
> Uviewer.plot()
> Vviewer.plot()
> Wviewer.plot()
> #
> toc = clock()
> print "Finished in %.3f secs."%(toc-tic)
>
>
>
>
--
Daniel Wheeler
ut how fipy handles it.
> I see there a reference to
> http://www.ctcms.nist.gov/fipy/documentation/refs.bib_cited.html#croftphd,
> but don't find access to that from my university.
Thanks. I'll delete that.
--
Daniel Wheeler
itories at some point. Feel
free to test it if you have an appropriate platform. Details for
building the deb and testing are here
<http://matforge.org/fipy/blog/FiPyDeb>. Cheers.
--
Daniel Wheeler
iscosity) + pressure.getGrad().dot([0.,1.])
>
> Would be nice if sombody can tweak it to have matching results with the
> literature, but anyway, this test has shown me fipy is adequate for my
> purposes.
> Obviously I wonder if the fully coupled matrix equation Daniel talks about
> would not also solve the problem better, however, that fortran code I refer
> to works apparently great without such a coupled solver.
>
> Benny
>
--
Daniel Wheeler
ield based on a velocity. Back solving using the advection equation
when you know the evolution of a field can give you the normal
component of a velocity (not all the components).
> to the velocities already found in examples/phase/simple, so
> that I know I am correctly solving for it.
Hope this helps. Cheers.
--
Daniel Wheeler
resolved.
Building from source with Python 3.1 should work without problems.
Enjoy,
Ralf
___
NumPy-Discussion mailing list
numpy-discuss...@scipy.org
http://mail.scipy.org/mailman/listinfo/numpy-discussion
--
Daniel Wheeler
d enthought version.
>
> Regards,
> Adya
>
> From: fipy@nist.gov [f...@nist.gov] on behalf of Daniel Wheeler
> [daniel.wheel...@gmail.com]
> Sent: Wednesday, July 28, 2010 9:41 AM
> To: Multiple recipients of list
> Subject: Re: Problem
On Wed, Jul 28, 2010 at 10:25 AM, Daniel Wheeler
wrote:
> * Try deleting everything under C:\Pythonxy and install Enthough's
That should be "C:\Python26".
--
Daniel Wheeler
; 2.When i try to import jdsym from pysparse i got this error
>
> Traceback (most recent call last):
> File "", line 1, in
> from pysparse import jdsym
> ImportError: DLL load failed: %1 is not a valid Win32 application.
>
> I've been trying to get this w
ou pleas help me resolve this issue.
>
> Thanks and Regards,
> Adya
>
--
Daniel Wheeler
ously I wonder if the fully coupled matrix equation Daniel talks about
> would not also solve the problem better, however, that fortran code I refer
> to works apparently great without such a coupled solver.
It will be interesting to see when when we have that functionality in fipy.
--
Daniel Wheeler
501 - 600 of 965 matches
Mail list logo
