CIS-Floquet

For finding eigenvalues of Floquet Hamiltonian with the Hamiltonian from a CIS basis

How to run:

python3 main.py path/to/input_file

How to use the fortran+MKL routines for block-LU factorization:

• Set the oneapi environment variables: source /path/to/intel/oneapi/setvars.sh

• Make sure that the compiler and linker flags are properly set in the Makefile and environment variables, use this link to get the correct flags: https://www.intel.com/content/www/us/en/developer/tools/oneapi/onemkl-link-line-advisor.html \

  • Select the appropriate MKL version
  • Select OpenMP threading (have not tried if TBB works)
  • Using dynamic linking is the easiest.
  • Make sure to select the appropriate interface layer, (check the size of integer variables in Fortran, e.g. sizeof(1)==4 means 32-bit integers)

• Add the suggested link line to the LDFLAGS environment variable eg:

  export LDFLAGS="-m64 -L${MKLROOT}/lib/intel64 -Wl,--no-as-needed -lmkl_gf_lp64 -lmkl_gnu_thread -lmkl_core -lgomp -lpthread -lm -ldl"

• Add the compiler options to the Makefile (if they differ from what's already in there)

• Compile factor class and interface the block_LU module with Python:

   cd src
   
   make block_LU
   
   cd ..
  

• Set use_fortran = t in the input file

  Special instructions for running on kestrel

  The MKL version installed on kestrel is outdated, and the above steps need to be modified slightly for things to work

  • Select the 2021 MKL version and Single Dynamic Library in the link line advisor
  • Add -shared to LDFLAGS (this is not related to mkl I think)
  • LDFLAGS should now look like this:

    -shared -m64  -L${MKLROOT}/lib/intel64 -lmkl_rt -Wl,--no-as-needed -lpthread -lm -ldl
    

  • Set the LD_PRELOAD environment variable:

    export LD_PRELOAD=/opt/intel/oneapi/mkl/latest/lib/intel64/libmkl_rt.so
    

  • Set the interface layer that the single dynamic library should use:

    export MKL_INTERFACE_LAYER=GNU,LP64
    

    See this link for more info: https://www.intel.com/content/www/us/en/docs/onemkl/developer-guide-linux/2024-2/dynamic-select-the-interface-and-threading-layer.html
  • Optional: change the threading layer. Default uses intel threading, but you can change to GNU with

    export MKL_THREADING_LAYER=GNU