By declaring Pd and Pu as CellVariables, you are declaring them to be functions
of space. That's what a CellVariable is.
Furthermore, eqpd and eqpu depend on Pf and Pk, so that necessitates that Pd
and Pu are functions of space.
Can you show us the math for the equations you are trying to solv
400,)
>
> 1 cellVariable (240,) (400,)
>
> 0 faceVariable (512,) (512,)
>
> 1 faceVariable (512,) (512,)
>
> 1 center coordinates (2, 240) (2, 400)
>
> 0 center coordinates (2, 240) (2, 400)
>
> 0 face coordinates (2, 512) (2, 512)
> 1 face coordinates
ument
> appears to be the right call after all.
>
> Working on the PR.
>
> Trevor
>
> ________
> From: fipy-boun...@nist.gov on behalf of Guyer,
> Jonathan E. Dr. (Fed)
> Sent: Wednesday, April 27, 2016 4:39:05 PM
> To: FIPY
>
gt; mpi4pyCommWrapper.py.
>
> Trevor
>
>
> ____
> From: fipy-boun...@nist.gov on behalf of Guyer,
> Jonathan E. Dr. (Fed)
> Sent: Wednesday, April 27, 2016 3:53:39 PM
> To: FIPY
> Subject: Re: globalValue in parallel
>
> It loo
It looks like 'recvobj' was removed from mpi4py about two years ago:
https://bitbucket.org/mpi4py/mpi4py/commits/3d8503a11d320dd1c3030ec0dbce95f63b0ba602
but I'm not sure when it made it into the released version.
It looks like you can safely edit fipy/tools/comms/mpi4pyCommWrapper.py to
rem
coeff=(phi+1))
>
> is like doing
>
> ImplicitSourceTerm(coeff=2)
>
> because I initially defined phi as
>
> phi = CellVariable(mesh=mesh, value=1)
>
> How should I define phi so that the source term is not a constant?
> ________
&g
-ImplicitSourceTerm(coeff=(phi+1))
will do it, although you may want to look at
http://www.ctcms.nist.gov/fipy/examples/phase/generated/examples.phase.simple.html
for a discussion of the most efficient way to linearize a source.
> On Apr 22, 2016, at 12:22 PM, Gaury, Benoit H. (IntlAssoc)
>
I don't see anything obviously wrong with the way you've posed the source. Why
do you think you have a problem?
> On Apr 18, 2016, at 4:44 PM, John Leeman wrote:
>
> Hi all,
>
> I’m working on implementing a rather simple thermal diffusion model with
> FiPy, but haven’t been able to find any
> On Apr 18, 2016, at 10:54 AM, Daniel Wheeler
> wrote:
>
> On Fri, Apr 15, 2016 at 9:29 AM, Phil wrote:
>> I see questions posted on this list, Gist and stack overflow - is there a
>> preference for posting ?
>
> I prefer Stackoverflow, but depends on the nature of the question.
> Stackove
PM, Raymond Smith wrote:
> Thanks, Jon.
>
> I'll post something to StackOverflow in the next few days.
>
> Also, that makes sense about the steady state ignoring the initial condition
> and finding the zero solution. I'd forgotten seeing that in the example.
>
>
Ray -
Thanks for your very complete and very correct answers to Dario.
If you wouldn't mind transcribing your answer to Dario's question on
StackOverflow, I'd be happy to up-vote it.
In answer to *your* questions:
> On Apr 3, 2016, at 11:09 AM, Raymond Smith wrote:
>
> Actually, I'm not su
sfied
> dependencies on my machine. So I must apologize, I must have been more tired
> than I thought.
>
> Sorry for the error!
> -Mike Waters
>
> On 3/30/16 11:52 AM, Guyer, Jonathan E. Dr. (Fed) wrote:
>> It looked to me like steps and accuracy were the way to do it,
ry that
> and see if my problem persists.
>
> Thanks again!
> -Mike Waters
>
> On 3/30/16 11:52 AM, Guyer, Jonathan E. Dr. (Fed) wrote:
>> It looked to me like steps and accuracy were the way to do it, but my runs
>> finish in one step, so I was confused. When I cha
Thanks,
-mike
On 3/29/16 3:38 PM, Guyer, Jonathan E. Dr. (Fed) wrote:
I guess I spoke too soon. FWIW, I'm running Trilinos version: 11.10.2.
On Mar 29, 2016, at 3:34 PM, Guyer, Jonathan E. Dr. (Fed)
mailto:jonathan.gu...@nist.gov>> wrote:
I'm not seeing a leak. The below is for tri
I guess I spoke too soon. FWIW, I'm running Trilinos version: 11.10.2.
On Mar 29, 2016, at 3:34 PM, Guyer, Jonathan E. Dr. (Fed)
wrote:
> I'm not seeing a leak. The below is for trilinos. VSIZE grows to about 11 MiB
> and saturates and RSS saturates at around 5 MiB. VSIZE
I'm not seeing a leak. The below is for trilinos. VSIZE grows to about 11 MiB
and saturates and RSS saturates at around 5 MiB. VSIZE is more relevant for
tracking leaks, as RSS is deeply tied to your system's swapping architecture
and what else is running; either way, neither seems to be leaking
On Mar 28, 2016, at 3:04 PM, Djoeke Schoonenberg
wrote:
> Hmm, when I do as you suggest:
>
> xface = mesh.faceCenters[0]
>
> Eq = TransientTerm(var=S_vap) == \
> ConvectionTerm(coeff=vgas(xface)+p*Dif/xface, var=S_vap)\
> + ImplicitDiffusionTerm(coeff=Dif, var=S_vap)\
>
> I
Mike, thanks for the example, and for the rather perverse application of FiPy!
I'll fiddle with this and see what I get.
- Jon
On Mar 25, 2016, at 7:16 PM, Michael Waters wrote:
> Hello,
>
> I still have a large memory leak when using Trilinos. I am not sure where to
> start looking so I mad
On Mar 28, 2016, at 12:20 PM, Djoeke Schoonenberg
wrote:
> So it seems that I’m implementing the variable speed in a wrong way. What I
> did was ConvectionTerm(FaceVariable(mesh=mesh, value=vgas(x)), var=S_vap),
> where vgas(x) is the speed as function of position. How should this be done
>
x27;s worth, here's what I was trying to run before. I think you'll see
the Gmsh code is the same as what is in your example
On Thu, Mar 24, 2016 at 3:53 PM, Guyer, Jonathan E. Dr. (Fed)
mailto:jonathan.gu...@nist.gov>> wrote:
OK, thank you.
Diffusion_Circle.py is not the n
an~1\appdata\local\temp\tmp2hf2nq.geo'
> Info: Meshing 1D...
> Info: Done meshing 1D (0 s)
> Info: Meshing 2D...
> Info: Done meshing 2D (0 s)
> Info: 1 vertices 1 elements
> Info: Writing 'c:\users\ddesan~1\appdata\local\temp\tmpisuemv.msh'
What platform are you running on?
Gmsh 2.0 is ancient and it's possible we've broken compatibility in some way
(although I don't think so). What happens if you type `gmsh --version` at the
command-line with one of the newer versions of Gmsh you've tried?
On Mar 24, 2016, at 11:56 AM, Daniel De
Thanks, Ray, for pointing to the instructions. I've unsubscribed Chris manually.
- Jon
On Mar 24, 2016, at 12:19 PM, Raymond Smith wrote:
> Chris, please try following the instructions on the website and putting
> unsubscribe in the body and sending the mail to fipy-requ...@nist.gov ...
>
> h
As Ray says, FiPy comes with a lot of examples that are intended to be an
intrinsic part of the documentation. Not all examples have been formatted and
included on the website or in the PDF (some don't have enough didactic, some
are redundant to other examples and are intended more as tests than
Interesting problem. Memory profiling has always been challenging, but I've not
seen signs of leaks when I've looked. Your PySparse w/ GC curve seems to
reinforce that. There's a lot more going on with Trilinos, but I don't know why
it would be leaky, even with garbage collection. Out of curiosi
r unrelated question: shall I preferentially make implicit my
> “reaction” term (= source term with opposite sign in two PDEs) ?
>
> Many thanks again.
> Best,
>
> Hervé
>
>> On Mar 10, 2016, at 17:49, Guyer, Jonathan E. Dr. (Fed)
>> wrote:
>>
>>&g
> I'm not entirely sure why the tuple version breaks, but the convection
> terms do store different variables.
A tuple of Variable fields is never a higher rank field. It might work
occasionally, but it really shouldn't be counted on.
I have no idea why V[None] produces a vector from a scalar, b
I assume there's more to the problem than this? With rho_h = 0, isn't the
initial condition phi_h = 0 already at the final solution?
On Mar 8, 2016, at 2:36 PM, Michael Waters wrote:
> I did some more testing, the trilinos LinearGMRESsolver works even
> better than the Pysparse LinearPCG and a
rgot to add proper units to b). Now it works normally. Sorry
> for the inconvenience. You may erase the thread if you wish.
>
> Regards,
> Gerard
>
>
>
> On 03/02/2016 09:29 PM, Guyer, Jonathan E. Dr. (Fed) wrote:
>> Please post the entire traceback. That one error line
Please post the entire traceback. That one error line is not enough to tell
what's going on.
On Mar 2, 2016, at 12:50 PM, Gerard Salvatella
wrote:
> Hello! I'm trying to solve a diffusion equation where the coefficient of
> the diffusion term depends on both x and the variable to solve
> (temp
es Pringle wrote:
> Thank you, Jon, that is helpful. I would love to have the ability to find the
> individual terms of the equation as calculated in the matrix that is solved
> by fiPy. This would make closing budgets, etc, easier. But your solution is a
> good start.
>
> T
James -
I don't think there's a straightforward way to get at this.
You can certainly write the explicit forms, e.g.,
(DiffCoeff * Psi.faceGrad).divergence
or
(Psi * convCoeff).divergence
but this isn't exactly the same as the matrix FiPy builds, as discussed at
https://github.com/usnist
; Djoeke
>
> On 29 Jan 2016, at 20:04, Guyer, Jonathan E. Dr.
> wrote:
>
>> Djoeke -
>>
>> The residual returned by sweep is the L2 norm of (L x - b) (see
>> https://github.com/usnistgov/fipy/blob/develop/fipy/solvers/solver.py#L143).
>> In other wor
I agree with everything Ray said. In electrochemical systems (which is what
Aniruddha was working on last I knew), swapping an arithmetic mean for a
harmonic mean makes an enormous difference. In semiconductors, there's an even
more non-linear interpolation that's widely used, known as Scharfett
play
> with this some more unless you have any ideas?
>
> I'll post updates just in case anyone else ever runs into these problems.
>
> On Mon, Feb 8, 2016 at 3:51 PM, Guyer, Jonathan E. Dr.
> wrote:
> Strange. I sort of understand what this error message means, but I don&
t; File
> "C:\Users\ddesantis\AppData\Local\Continuum\Anaconda2\lib\site-packages\fipy\variables\histogramVariable.py",
> line 38, in
> from fipy.meshes import Grid1D
>
> File
> "C:\Users\ddesantis\AppData\Local\Continuum\Anaconda2\lib\site-packages\fipy\mes
from fipy.variables.histogramVariable import *
>
> File
> "C:\Users\ddesantis\AppData\Local\Continuum\Anaconda2\lib\site-packages\fipy\variables\histogramVariable.py",
> line 38, in
> from fipy.meshes import Grid1D
>
> File
> "C:\Users\ddesan
I would guess that you've run into the problem in
https://github.com/usnistgov/fipy/issues/462
We fixed this last summer, but haven't pushed a new release since then (sorry
about that).
If you check out the `develop` branch from our repository, I think that will
correct this problem.
If not,
For an individual run, append --trilinos to the command line. To set it for all
runs going forward, set the environment variable (bash syntax):
export FIPY_SOLVERS=trilinos
See
http://www.ctcms.nist.gov/fipy/documentation/USAGE.html#command-line-flags-and-environment-variables
On Feb 1, 201
Djoeke -
The residual returned by sweep is the L2 norm of (L x - b) (see
https://github.com/usnistgov/fipy/blob/develop/fipy/solvers/solver.py#L143). In
other words, it is the unscaled deviation from zero of the linear
discretization of your non-linear PDEs. Because we don't scale the residuals
There should be no reason for incompatibility between 1 and var2.
max doesn't know about numpy arrays or FiPy Variables.
numerix.fmax should work, but doesn't. Oddly, numerix.fmax((1-var2), a) does
work. I vaguely know why, but need to delve deeper
(https://github.com/usnistgov/fipy/issues/486)
do
> each sweep increases with each time step, independent of the time step size.
> Could this be a memory issue?
>
> ~Dan
>
> On Mon, Dec 21, 2015 at 10:42 AM, Guyer, Jonathan E. Dr.
> wrote:
> Dan -
>
> I recommend you plot the residual as a function of sweeps fo
Be sure to start with a clean build directory after changing any settings. I
would guess that -DTPL_ENABLE_Boost=OFF didn't work because traces of the
previous cmake run will still present.
Our instructions are almost certainly out of date. I find building Trilinos
unpleasant and it's been almo
r. It starts from ~0.5s and increases by ~0.01s with each
> sweep, quickly growing to levels too slow to both simulating. Any guidence on
> this would be appreciated.
>
> ~Dan
>
> On Sat, Oct 31, 2015 at 2:04 PM, Guyer, Jonathan E. Dr.
> wrote:
> Glad to hear it.
>
3.3.
On Dec 4, 2015, at 5:27 PM, Guyer, Jonathan E. Dr.
wrote:
> Survey closes shortly. Vote early and vote often.
>
> On Dec 1, 2015, at 9:20 AM, Guyer, Jonathan E. Dr.
> wrote:
>
>> Survey results are overwhelming, but statistically insignificant.
>>
>>
Survey closes shortly. Vote early and vote often.
On Dec 1, 2015, at 9:20 AM, Guyer, Jonathan E. Dr.
wrote:
> Survey results are overwhelming, but statistically insignificant.
>
> We will close the survey on Friday, December 4th, 2359 GMT and determine our
> path forward a
Guyer, Jonathan E. Dr.
wrote:
> [tl;dr version: please take the survey at
> https://www.surveymonkey.com/r/KVYHFKK]
>
> According to our documentation (not that we've tested in awhile) FiPy
> currently supports all Python versions from 2.4 on. If you use the 2to3
> utility,
[tl;dr version: please take the survey at
https://www.surveymonkey.com/r/KVYHFKK]
According to our documentation (not that we've tested in awhile) FiPy currently
supports all Python versions from 2.4 on. If you use the 2to3 utility, FiPy
also works with Python 3, but this isn't a configuration
; Yun
>
>
>
> On Tue, Oct 27, 2015 at 4:55 PM, Yun Tao wrote:
> Aha! That solved it!! Thanks so much, Jon! I'll be sure to look into the
> subclass declaration of `CellVariable'.
>
> Cheers,
> Yun
>
> On Tue, Oct 27, 2015 at 4:46 PM, Guyer, Jonathan E.
Try phi.grad.dot(phi.grad)
On: 04 November 2015 21:28, "Fausto Arinos de A. Barbuto"
wrote:
>T.grad()*T.grad()
I mean, phi.grad()*phi.grad(). Yet another typo.
Fausto
___
fipy mailing list
fipy@nist.gov
http://www.
That term is really the partial derivative of \phi with respect to the quantity
(x^2)? Strange. I don't think you have any choice but to do that piece with the
chain rule. After that, you'll get
(\frac{1}{2 x}\frac{\partial \phi}{\partial x})^2
but that's still not a form that FiPy can really d
SIMPLE scheme if
> I should run into any more trouble.
>
> Best regards,
> Dion
>
> On 31.10.2015 19:03, Guyer, Jonathan E. Dr. wrote:
>> Dion -
>>
>> Thanks for your example and explanation. I've converted to an IPython
>> notebook to make it
e dependence. I agree that
> derivations of the heat equation are often not careful about non-linear
> coefficients. I'll try your reformulation of it later this week. We have some
> other tools in-house to benchmark it with.
>
> ~Dan
>
> On Wed, Oct 28, 2015 at 5
Dion -
Thanks for your example and explanation. I've converted to an IPython notebook
to make it easier to show the math derivation. See
https://github.com/guyer/haefner/blob/master/killworth.ipynb (math rendering is
kind of random on the github site, but should be OK if you download it and hav
Dan -
Interesting problem. What I actually find is that the residual decays or at
least holds fairly steady until some point at which the residual rises rapidly,
regardless of the number of sweeps.
I was ultimately able to make some headway by looking at the form of your
equations. By writing
In 3D this is equivalent to
> restricting the gradient to be perpendicular to the conductor surface. Does
> FiPy allow applying a boundary condition like this for an interior object?
>
> Kevin
>
>
>
> From: Guyer, Jonathan E. Dr.
> Sent: Thursday, October 22, 201
e of an embedded function update_signal inside the iteration loop
> messing things up?
>
> Thanks,
> Yun
>
> On Mon, Oct 19, 2015 at 4:35 PM, Guyer, Jonathan E. Dr.
> wrote:
> You would do this in a similar way to how time-dependent boundary conditions
> are illustrated
compatible on the cluster here. Hopefully a new version of mpi4py will
> be installed soon so I may continue to parallelize.
>
> Thank you,
> -mike waters
>
> On 10/22/2015 11:50 AM, Guyer, Jonathan E. Dr. wrote:
>> Mike -
>>
>> The most expedient way to get th
carefully, and I'm happy to share it if
> anyone is curious about the aspects of this problem that may not make it a
> great fit for fipy. (Though I'm definitely not familiar with numerical PDE
> solvers in general so I'm sure it's mostly my lack of facility with th
Mike -
The most expedient way to get this normalization is
rho / (rho.cellVolumeAverage * mesh.cellVolumes.sum())
`cellVolumeAverage` and `cellVolumes.sum()` automatically take care of the MPI
communication and account for not double-counting the overlaps between
partitions.
Also, if you us
Peter -
I'm glad sweeping was able to get you a more reasonable solution (although I'm
not seeing the non-linearity that would require it). `AdvectionTerm` only ever
gets used in our level set examples, which I'm not intimately familiar with. I
don't know of any fundamental reason it wouldn't w
The discussion at
http://www.ctcms.nist.gov/fipy/documentation/USAGE.html#applying-internal-boundary-conditions
describes exactly what you are trying to do. I think I know why it might not
have worked for you, though. When declaring an `ImplicitSourceTerm`, FiPy has
to be careful not to add neg
, for some reason, var(x,t) is unresponsive to these
> changes.
>
> Therefore, my question is: how can I base the convection term on a
> CellVariable that gets temporally updated outside of the equation definition?
>
> Thanks,
> Yun
>
> On Wed, Aug 26, 2015 at 2:47 PM, G
1 2
> , from bottom row to top row, left to right column, when I call the viewer?
> Sorry for these simple questions, I just started using fipy
> Thanks
> Kevin
>
> On Wed, Oct 7, 2015 at 1:56 PM, Guyer, Jonathan E. Dr.
> wrote:
> Please see:
>
> exampl
Please see:
examples/phase/binary.py
examples/phase/polyxtalCoupled.py
examples/phase/quaternary.py
examples/reactiveWetting/liquidVapor2D.py
for some examples of spatially varying convection velocity fields.
On Oct 6, 2015, at 6:54 PM, Kevin Lin wrote:
> Hi all,
>
> I am trying to use fipy t
Jose -
What's surprising to me is how infrequently this question has come up. I don't
think there's anything wrong with what you've done, but understand your feeling
that you've "hacked" FiPy. It's quite possible that you can just write `T =
bellLinT(Sigma)`. Depending on what's in bellLinT, th
r recommendations and I'm able to get a solution. Now I need to just need
> to troubleshoot the model so that I can get the _correct_ solution.
>
> Again, I really appreciate your thoughtfulness and recommendations!
>
> Kind regards,
>
> Douwe Bruinsma
>
> On Fr
We used to have the ability to multiply a term by another field, but this
capability was removed some time ago because it didn't work properly a lot of
the time.
Even if you could do it, I don't think it would help. Instead of dividing by
zero on the b-vector, you'd be multiplying be zero on th
Your thoughts make sense and you're not missing an API.
As I understand it, I think extracting the flux using the weights from the
numerical scheme is the most accurate representation of the physical flux, as
these account for the likely fact that the field does not vary linearly between
cell c
r both method, but I think my Robin boundary condition
implementation in FIPy is wrong. So I would be appreciated if you look at my
equation and code and help me to correct the boundary condition definition.
Thanks,
Vahid
On Monday, September 28, 2015 3:17 PM, "Guyer, Jonathan E. Dr."
What was the result you were expecting?
As discussed in the email thread you linked to, (1) has a known bug:
https://github.com/usnistgov/fipy/issues/426
I also discovered that the typeset analytical expression in the robin.py
example was also wrong, although the analytical solution in the code
Vahid -
The following allows you to access the matrix after solving or sweeping:
>>> myEquation.cacheMatrix()
>>> myEquation.solve(...)
>>> matrix = myEquation.matrix
The exact form of the matrix will vary depending on the solver suite you use
(PySparse, PyTrilinos, ...).
Similarly, you can ac
Zebo -
A few points to augment Trevor's (correct) answer:
* I think I know what mean when you distinguish between "discretized region"
and "continuum region", but be aware that FiPy does not solve continuum
equations. Any FiPy mesh is a discretization in space to enable solution by the
finite
You'll need to give us a bit more to work with. As much as possible, FiPy
Variable's just work with numpy. Please use our fipy.tools.numerix library
instead of importing numpy directly, as we patch a few things, but largely you
can just pass our Variables to numpy and get Variables back. If you
Douwe -
Sorry for the delay in answering. I had to think through this a bit.
The most immediate cause of the error you received is that you are attempting
to solve for u declared as a FaceVariable. FiPy can only solve equations for
CellVariables. This is easily remedied by changing the declarat
A large number of methods and classes have been deprecated in FiPy for a long
time and it's time to clean up the code. The 3.2 release of FiPy (hopefully
coming soon) will be removing anything that was deprecated in FiPy 3.0 or
earlier. This includes just about any method that starts with "get"
Yun -
I've gotten your script to "work" and posted the changes to:
https://gist.github.com/guyer/caca956463dfc3835722/revisions
The main changes I made were:
* to get rid of the intrep2d, as it wasn't working properly [signal_fct(xf, yf)
generates a result of shape
(len(xf), len(xf)) inst
Thanks, Trevor, for continuing to plug at this, and for sorting out the
threading issue.
FYI, your efforts to find pygist are commendable, but ultimately futile.
https://github.com/usnistgov/fipy/pull/453, which you accepted and merged, did
away with the GistViewers, as they were deprecated yea
Try the solutions at:
http://permalink.gmane.org/gmane.comp.python.fipy/1576
and
http://stackoverflow.com/a/29996927/2019542
On Jun 23, 2015, at 8:57 AM, fomin wrote:
> I solve an equation for a three-dimensional grid. How do I get 2-D
> cross-section of the solution of the equation of
Since neither of the primary sparse solvers we use, PySparse or PyTrilinos, has
been migrated to Python 3, we're not in any rush.
FiPy can be used with Python 3 by running the `2to3` converter and then using
the SciPy sparse solvers. This doesn't perform as well, but is suitable if you
need to
You can certainly write f.grad.dot(g.grad), but that will be fully explicit and
will not converge very well.
It's best to look at the equation in its entirety. Where did this term come
from? Oftentimes, terms like this arise from relations like
\nabla\cdot(f\nabla g) \equiv \nabla f \cdot \n
I'm glad to hear that worked for you.
On Mar 21, 2015, at 7:36 PM, abart...@caltech.edu wrote:
> I made the changes that you suggested. The program works perfectly now.
> Thank you for the help. This should shave hours off of my simulations.
>
> Thanks again,
> Tony
>
>
>> It doesn't have anyt
It doesn't have anything to do with the solver.
The issue is that multiprocessing pickles the arguments to the multiprocessed
function. FiPy has known issues with pickling (see
https://github.com/usnistgov/fipy/issues/358 and
https://github.com/usnistgov/fipy/wiki/MeshPickling#sour-pickles (wh
On Mar 18, 2015, at 10:06 AM, "Daniel Wheeler"
wrote:
>> I would appreciate contact to anyone who has developer experience with FiPy
>> to discuss my implentations to keep them in line of the structure of the
>> FiPy project/code.
>
> Let's set up a video chat with Jon Guyer included. Email m
I'm not aware that we've encountered this, but I also don't think we've ever
tried.
Just to clarify, the issue is that the answer changes between processes=1 or
processes=2+?
I don't think that any of FiPy's parallel code (e.g., using mpirun to invoke
PyTrilinos in parallel) should be triggere
y project. It is
> complicated one but I tried on Linux and it works fine. Let me try to import
> fipy project and run it through eclipse and also will try to run on cmd
> prompt.
>
>
> On Friday, March 13, 2015 10:53 PM, "Guyer, Jonathan E. Dr."
> wrote:
>
&g
I suspect that `which pip` will produce /usr/local/bin/pip and thus will be
associated with /usr/local/bin/python. For whatever reason your $PATH finds
/usr/bin/python before /usr/local/bin/python.
pip and fipy (and all of fipy's prerequisites) all need to be installed into
the same python inst
he output is too long. I am attaching the last part of the output. What
> about installing fipy on Win7. Is it still a problem?
>
>> On 3/12/2015 3:42 AM, Guyer, Jonathan E. Dr. wrote:
>> Dhisa -
>>
>> We're not allowed to run Windows 8 here, so we have
Dhisa -
We're not allowed to run Windows 8 here, so we have no experience with this.
Can you provide the output of
>>> import fipy
>>> fipy.test()
On Mar 10, 2015, at 4:44 AM, Dhisa Minerva wrote:
> I tried to install fipy and it's dependencies on 2 windows PC and the graph
> doesn't sh
I am no specialist in noise terms, either, but the \delta(\vec{r} -
\vec{r}')\delta(t - t') terms express "uncorrelated in space and time", so both
\xi_c and \xi_\eta have that property in what you attached.
Equation (1a) with the Laplacian is the appropriate conservative form for the
noise on
cmake \
-D CMAKE_BUILD_TYPE:STRING=RELEASE \
-D Trilinos_ENABLE_PyTrilinos:BOOL=ON \
-D BUILD_SHARED_LIBS:BOOL=ON \
-D Trilinos_ENABLE_ALL_OPTIONAL_PACKAGES:BOOL=ON \
-D TPL_ENABLE_MPI:BOOL=ON \
-D MPI_BASE_DIR:PATH=${ANACONDA} \
-D Trilinos_ENABLE_TESTS:BOOL=ON \
-D DART_TESTING_TI
On Feb 21, 2015, at 4:40 PM, Dhisa Minerva wrote:
> I am having a problem here in defining the variable name if using vector
> form. For example I have 3 components, v1, v2, and v3. I define them in the
> vector form as v. I want to put the name v1_sol, v2_sol, and v3_sol for
> variable v1, v
Let me add that if you performed such an analysis, that would be a welcome
addition to the example, especially if it shows that your decomposition scheme
is superior.
On Feb 23, 2015, at 8:44 AM, Guyer, Jonathan E. Dr.
wrote:
> I don't know. The form you write down in "split pl
oeff=epsilon**2, var=phi))
>
>
> eq = eq1 & eq2
> dt = 0.5
> viewer = fp.Viewer(vars=(phi, psi))
> for t in range(1000):
> phi.updateOld()
> psi.updateOld()
> res = 1.e5
> while res > 1.e-1:
> res = eq.sweep(dt=dt)
> print res
> viewer.plot()
On Feb 20, 2015, at 12:08 PM, Daniel Wheeler wrote:
> On Fri, Feb 20, 2015 at 6:39 AM, Ronghai Wu wrote:
>> Thanks, I worked it out. But still have three questions:
>> (1) The 4th-order is split by "psi = d2fdphi2(phi-phiold) + dfdphi -
>> epsilon**2*laplace phi". I do not understand why we nee
On Feb 20, 2015, at 6:39 AM, Ronghai Wu wrote:
> (1) The 4th-order is split by "psi = d2fdphi2(phi-phiold) + dfdphi -
> epsilon**2*laplace phi". I do not understand why we need
> d2fdphi2(phi-phiold)?
See the discussion on linearization of the source in
http://www.ctcms.nist.gov/fipy/exampl
Those figures were just copied over from
http://www.ctcms.nist.gov/fipy/examples/cahnHilliard/generated/examples.cahnHilliard.mesh2D.html
and represent a 1000x1000 mesh run to an elapsed time of 1000.
The coupled example runs a 20x20 mesh up to t=0.05. You should be able to
manually increase t
I understand what you mean about it being simpler, but I tend not to work in
1D, because it can be misleading. You've got combinations of even-order and
odd-order terms that don't make much sense to me if you think about them in
higher dimensions. Using \nabla instead of (\partial / \partial x)
Without knowing what the errors are, it's hard to be sure.
Do FiPy's examples seem to run properly?
On Feb 6, 2015, at 7:49 PM, Kyle Lawlor wrote:
> Thanks for the replies.
> I ended up using the pip install method.
> Which seems to have worked pretty well up to 4 errors (after running
> fip
isn't it? Or is it
> because the domain solution is 1D? but it doesn't make sense since what we
> have here is 2 variables.
>
> Look forward to your reply
>
> Dhisa
> On Friday, February 6, 2015 8:08 AM, "Guyer, Jonathan E. Dr."
> wrote:
>
>
201 - 300 of 412 matches
Mail list logo