DAMPyF: a Python implementation of the DAMPF method for the simulation of open-system dynamics

DAMPyF is a Python software package designed to simulate the dynamics of open quantum systems. It implements the dissipation-assisted matrix product factorization (DAMPF) method, which is particularly suited for systems where the environment exhibits memory effects (non-Markovianity) and structured spectral densities. By representing the environment through a set of damped harmonic oscillators—known as pseudomodes—and utilizing tensor-network techniques, the software enables numerically exact simulations of complex quantum dynamics that would otherwise be computationally prohibitive.

The core of the DAMPF approach is the decomposition of the extended system's density matrix into a matrix-product-state (MPS) representation. Unlike traditional methods that might struggle with the exponential growth of the Hilbert space, DAMPF keeps the system indices explicit while encoding the pseudomode degrees of freedom into MPSs. This structure is highly efficient because the local dissipation inherent in the pseudomode model suppresses the growth of correlations, allowing for manageable bond dimensions even in systems with many environmental modes. The time evolution is performed using a Trotter decomposition of the Liouvillian, which breaks the dynamics into manageable steps involving the system, pseudomodes, and their coupling.

DAMPyF provides two primary workflows to accommodate common research needs in quantum physics and chemistry:

1. Excitation Energy-Transfer Dynamics: This mode propagates a system excitation in time, allowing researchers to study how energy moves through multi-site systems.
2. Molecular Spectroscopy: This high-level workflow is tailored for computing linear spectra, such as absorption and circular dichroism, by propagating optical coherences within electronic systems.

The software is designed for reproducibility, utilizing a centralized configuration file and standardized text-based input formats for Hamiltonians and pseudomode parameters. It includes helper tools for generating input data, analyzing results, and deploying simulations on high-performance computing clusters.