Pydiso is a python wrapper for the pardiso solver. It is currently implemented for the Intel MKL's version of pardiso. It's goal is to expose the full functionality of pardiso to python, allowing the user to customize it as necessary for their use cases.
The wrapper is written in cython and links to the mkl libraries dynamically. Therefore, it needs to find the necessary header files associated with the MKL installation to compile. You will also need to clone this repository.
For conda users, these headers can be installed with mkl-devel package that is available
on the default channel, conda-forge channel, the intel channel, or others, e.g.
conda install mkl-devel
If you are not using a conda environment, you can install the MKL header files with:
apt-get update && apt-get install intel-mkl
If you installed with conda, it is likely that your respective numpy installations will then
identify the location of the installed MKL. However if you run into issues automatically
finding the library headers, you will need to set the MKLROOT environment variable to
point to the correct location.
The mkl-rt library and mkl.h may be in the same folder. In that case MKLROOT should point to
that directory. If you installed mkl with apt-get, it is likely your directory structure will
look something like this:
usr/
include/
mkl/
mkl.h
...
lib/
x86_64-linux-gnu/
libmkl_rt.so
...
In this case, set MKLROOT to the usr/ directory (the directory containing include and lib),
e.g. export MKLROOT="/usr".
The mkl-rt.h and mkl-rt.lib are not in the same folder but seperated a level into
Library and Include directories, and in this case MKLROOT would point to the folder
containing them.
After the necessary MKL files are accessible, you should be able to do the standard install
script common to python packages. Navigate to the pydiso directory, then run either
python setup.py install
or, equivalently
pip install ..