lindblad

Quantum optical master equation simulations.

The lindblad library provides functionality for numerical simulations with open quantum systems. The focus of the library is on medium sized problems with on the order of tens of quantum levels.

Example:

#include <Lindblad.hpp>
#include <vector>

static const int numIters = 1000;

int main() {
  MasterEqn meqn(2);
  double g = 1.0;
  meqn.addCoupling(0, 1, g);
  std::vector<Amplitude> rhoInitial(4, 0);
  rhoInitial[0] = 1.0;
  MasterEqnEvolution evolution(meqn, &rhoInitial[0]);
  for (int i = 0; i < numIters; ++i) {
    evolution.takeStep();
  }
  return 0;
}

Obtaining lindblad

The lindblad library is distributed in source form. It can be downloaded from the lindblad github repository. Configuration is via cmake. On linux the library is downloaded, configured, built, and installed with the following commands:

git clone https://github.com/d-meiser/lindblad
cd lindblad
mkdir build
cd build
cmake ..
make -j4
make install

Building PETSc

Several of the examples require the PETSc library. To build PETSc, blas and lapack headers and libraries are needed. A minimal configuration of PETSc can be created with the script ./utilities/get_petsc.sh. On ubuntu, the following works for me:

sudo apt-get install libblas-dev libatlas-dev liblapack-dev
cd $LINDBLAD_ROOT_DIRECTORY
sh ./utilities/get_petsc.sh

$LINDBLAD_ROOT_DIRECTORY is the directory in which this README file lives. The get_petsc.sh script builds a minimal PETSc library and installs it into the ${LINDBLAD_ROOT_DIRECTORY}/petsc-cpucmplx directory.

Once PETSc is built and installed lindblad can be configured with PETSc support as follows:

cd ${LINDBLAD_ROOT_DIRECTORY}/build
cmake -DPETSC_DIR=../petsc-cpucmplx ..
make

Overview of the source code

• include: Public API of the lindblad library
• examples: Some examples.
• include/detail: Internal headers.
• tests: Tests for lindblad.
• src: Implementation files of the library.

Questions, comments, and issues

Questions, comments, and issues should be reported via the github issue tracker.