This repository contains conda recipes for the FEniCS libraries. The recipes work with conda-build, the build system created for the Anaconda Python distribution.
Note: These recipes are no longer maintained and outstanding bugs are no longer being fixed. Please check out any of the project forks and this mailing list thread for more information.
Packages are generated for linux-64 in the juanlu001 binstar channel. To install them, just type:
$ conda install fenics --channel juanlu001
Note: These binaries are portable across Linux distributions but are compiled without VTK support. If you want plotting capabilities, please consider local building.
The recipes allow for building the packages locally, for later fast installation or even redistribution.
$ conda install conda-build
$ conda build eigen3 swig petsc petsc4py slepc instant ufl fiat ffc dolfin fenics --python 2.7
$ conda install fenics --use-local
Requirements:
- Bash (Anaconda only works with this shell)
You probably forgot to specify the channel, or to add them to your conda configuration, as explained above.
Maybe you installed the latest FEniCS version but your system is too old. Try to install FEniCS 1.4.0 from the channel https://conda.binstar.org/juanlu001/channel/fenics:1.4.0:centos as explained above.
There is some sort of linking problem in your system. Perhaps you have
to install some of the Boost system requirements, specially libbz2. To diagnose
this problem you can use the ldd utility:
ldd <INSTALL_PATH>/anaconda/envs/fenics27/lib/python2.7/site-packages/dolfin/cpp/_common.so
DOLFIN runtime dependency is not met. Install the following python module: 'ffc' and make sure its location is listed in PYTHONPATH
Many times this error has nothing to do with FFC being not installed. Please try
>>> import ffc
in a Python session to see the real problem.
Please open an issue at https://github.com/juanlu001/fenics-recipes/issues. In the meanwhile, you can try and build your own conda packages from the recipes.
- Anaconda Python distribution https://store.continuum.io/cshop/anaconda
- Introduction of conda-build http://continuum.io/blog/new-advances-in-conda
- Documentation http://conda.pydata.org/docs/build.html