Provide reliable, accurate down to several ULP's and fast numerical framework to compute generalized Fermi-Dirac integrals aiming at full floating-point parameter coverage, generality, precision control and speed.
To install library, follow standard Linux autoconf procedure (see also INSTALL):
./configure
make
sudo make install
sudo ldconfig
To install in e.g. home directory:
./configure --prefix=$HOME
make
make install
If you prefer to use other compilers, e.g:
./configure CC=icc CFLAGS="-qopenmp -xhost -O3"
Use of the library is straightforward in C and Fortran. See examples/ subdir, files example.c and example.f90, respectively.
To use library in Mathematica, first compile MathLink code (replace 12.3 with your Mathematica $VersionNumber), which is located in examples subdirectory (Fermi-Dirac.tm and Fermi-Dirac.c files).
/usr/local/Wolfram/Mathematica/12.3/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions/mprep Fermi-Dirac.tm -o Fermi-Dirac.tm.c
gcc -c Fermi-Dirac.tm.c Fermi-Dirac.c -I /usr/local/Wolfram/Mathematica/12.3/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions/
g++ Fermi-Dirac.tm.o Fermi-Dirac.o -o Fermi-Dirac -L /usr/local/Wolfram/Mathematica/12.3/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions/ -lML64i4 -lpthread -lrt -lstdc++ -ldl -luuid -lfermidirac
g++ Fermi-Dirac.tm.o Fermi-Dirac.o /usr/local/Wolfram/Mathematica/12.3/SystemFiles/Links/MathLink/DeveloperKit/Linux-x86-64/CompilerAdditions/libML64i4.a -o Fermi-Dirac -lpthread -lrt -lstdc++ -ldl -luuid -lfermidirac
You might need to install libuuid developement files. On Ubuntu: sudo apt install uuid-dev On RedHat: yum install libuuid-devel
System might not find Mathematica libraries, producing error. In this case run:
export LD_LIBRARY_PATH=$LD_LIBRARY_PATH:/usr/local/Wolfram/Mathematica/12.3/SystemFiles/Libraries/Linux-x86-64/
Then, in Mathematica use (or whatever directory it is):
SetDirectory["$HOME/libfermidirac-0.26/examples"]
Install["Fermi-Dirac"]
?FFermi
In:= FFermi[1.0, 1.0, 1.0]
Out = 2.64283
In:= GFermi[1.0, 1.0, 1.0]
Out = 58.676
It is assumed python3 is being used.
First, install cffi library
pip install requirements.txt
or simply
pip install cffi
Then, enter examples/ subdir and compile fast Python library using C sources:
python3 create_interface.py
Above step should create fermidirac.cpython-38-x86_64-linux-gnu.so library. Then, check the example:
python3 example.py