You cannot post messages because only members can post, and you are not currently a member.
Description:
CLAWPACK Developers
|
|
|
PyClaw's Controller class loses Fortran order of frames
|
| |
All,
I've filing a bug against this here:
[link]
It's enough of a gotcha that I thought I would send a general warning to
PyClaw developers and users.
The following clawpack controller code 'loses' the imposed Fortran order
passed in from the state object when a frame gets passed in:... more »
|
|
|
Fwd: DA memory allocation
|
| |
---------- Forwarded message ----------
...To: Lisandro Dalcin <dalc...@gmail.com>
Cc: david ketcheson <david.ketche...@kaust.edu.sa>
Hi Lisandro
This is a further analysis of the memory issue in PetClaw that Grady
noticed and discussed here:
[link]... more »
|
|
|
Option to not save previous q for Strang splitting in 3D PyClaw
|
| |
Hi all,
A notion just occurred to me regarding reducing memory usage. With Strang
splitting, all variants of Clawpack currently save the previous time step's
q array before advancing by the first half time step of the source term, in
case the hyperbolic step has to be rejected. For some problems (I'm... more »
|
|
|
3D PetClaw
|
| |
That sounds good, although it's not completely clear what the right way
would be to do this. I can see two obvious approaches:
1. The Clawpack Classic way -- create the qbc and auxbc arrays before the
source term is called, and pass the whole array into the source routine for
Strang splitting. This would be the most flexible approach, but might... more »
|
|
|
Test
|
| |
Can I post to this group?
|
|
|
Separating entropy fixes from Riemann solvers
|
| |
Gorune Ohannessian, who is developing CUDAClaw, brought to me a suggestion
today regarding the implementation of entropy fixes. The suggestion has
both aesthetic and performance consequences.
The computation of fluctuations (amdq, apdq) can be done in a generic way
independent of details of the Riemann solver. Thus it makes sense to do... more »
|
|
|
Clawpack Developers Hackathon July 22-26, 2013
|
| |
Hi all,
Several of us are planning to meet in Seattle the week of July 22,
2013 for some intensive Clawpack development. In particular the
following folks plan to be here for at least part of the week: Marsha
Berger, Kyle Mandli, David Ketcheson, and Donna Calhoun.
This will be loosely organized since different subsets of people are... more »
|
|
|
new "devel" branch in clawpack/clawpack
|
| |
All,
I've pushed a new branch named 'devel' to top-level clawpack/clawpack.
I'd like to use this branch for unstable development, that is, to keep
track of integrated but untested changes between all four repositories, as
opposed to bumps to clawpack/clawpack's master branch, which we have been... more »
|
|
|
How do we populate problem_data?
|
| |
All,
I know that we are still thinking about the Riemann functionality, and
whether it makes sense to move to heavier Fortran-90 modules or leave them
as light-weight functions. This is a separate issue with regards to how we
compose the problem_data object, especially when restarting/continuing an... more »
|
|
|
Riemann refactor for Clawpack 5.0
|
| |
I'm moving this conversation to claw-dev from
[link]
See that page first to catch up on the discussion.
Grady was wondering if putting the Riemann solver in a module and/or
using derived types causes any problems for PyClaw since f2py may have
problems with these?... more »
|
|
|
