pyapes
is designed to solve various engineering problems in rectangular grid.
The goal of pyapes
(should be/have) is
torch.Tensor
. User can choose either torch.device("cpu")
or torch.device("cuda")
.We recommend to use poetry
to manage/install all dependencies.
From git
git clone git@gitlab.ethz.ch:kchung/pyapes.git
cd pyapes
poetry install
From pypi
python3 -m pip install pyapes
# or
poetry add pyapes
python >= 3.10
torch
does not support 3.11 properly (for the official release). Therefore, stick to python3.10
.torch >= 1.10.0
pymyplot
(plotting tools)pymytools
(misc. tools including data I/O, logging, etc.)CPU/GPU(CUDA) computation using torch
(OpenFOAM like) generically expressed solver
>>> solver.set_eq(fdm.laplacian(1.0, var) == rhs)
>>> solver.solve()
FDM Discretizations
Grad
, Laplacian
, Div
upwind
for the Div
operatorDdt
Boundary conditions:
Dirichlet
, Neumann
, Periodic
, and Symmetry
Demo cases in jupter
notebooks
Check our demos files
quick
)Ddt
class (implementation is tested but haven't validated with practical test cases)