Skip to content

Instantly share code, notes, and snippets.

@wgwz
Last active August 29, 2015 14:27
Show Gist options
  • Save wgwz/2ac820f959666c9f5cd6 to your computer and use it in GitHub Desktop.
Save wgwz/2ac820f959666c9f5cd6 to your computer and use it in GitHub Desktop.
my installation of pymks

OS: Linux Mint 17.2

Using Anaconda Distribution Python


The following worked for installation of pymks.

This walks through setting up sfepy as well

Necessary libaries

sudo apt-get install gfortan gcc
sudo apt-get install liblapack-dev

Conda installs

conda install cython pysparsing pytables mayavi mpi4py

Pip installs

For igakit

pip install https://bitbucket.org/dalcinl/igakit/get/default.tar.gz

For metis

pip install metis

For scikit-umfpack

pip install scikit-umfpack

setup.py installs

For pysparse Download source code (in this case v. 1.1.1)

cd pysparse-1.1.1
python setup.py install

Checking pysparse install

cd Test
python test_superlu.py
python testSuperlu.py

Both should all pass and read ok

Finally sfepy as per pymks install procedure

git clone git://github.com/sfepy/sfepy.git
cd sfepy
python setup.py install

Final output will show:

cython is available in version 0.23
numpy is available in version 1.9.2
scipy is available in version 0.15.1
matplotlib is available in version 1.4.3
pyparsing is available in version 2.0.3
tables is available in version 3.2.0
mayavi (optional) is available in version 4.4.0
sympy (optional) is available in version 0.7.6
igakit (optional) is available in version 0.1
petsc4py (optional) is missing; you may get run-time errors
mpi4py (optional) is available in version 1.3.1
pymetis (optional) is available in version 2014.1
scikits.umfpack (optional) is available in version 0.1
pysparse (optional) is available in version 1.1.1

Note: I could not set up petsc4py

Now install pymks

pip install pymks

Rebuild scipy

conda install scipy

(in order to get umf compatible)

To check the install

python -c "import pymks; pymks.test()"

Output should be:

.....................................................................
----------------------------------------------------------------------
Ran 69 tests in 41.107s
OK
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment