Related work

What qscat shares with code that is already published, and what appears to have no released counterpart. Surveyed 2026-08-16 across PyPI, GitHub, Zenodo, and the published literature; the survey’s method and its blind spots are stated at the end, so a reader can judge how much the “no counterpart found” claims are worth.

Summary

Layer	Status
FEM-DVR grid, Gauss–Lobatto quadrature, bridge functions, ECS contour, Crank–Nicolson propagation	Also implemented elsewhere — quantumgrid (Python, MIT), plus several partial implementations in Fortran/Julia
Driven Lippmann–Schwinger solve, VE/DA/DR cross sections, LCP approximation, resonance levels, TD extractors, grid tuner	No released implementation found in any language
Electron–molecule scattering as a research goal	Well served by ab initio suites (UKRmol+, ePolyScat, FERM3D) that solve a different problem with a different method

The method qscat implements is not new — it is Rescigno & McCurdy’s FEM-DVR-ECS (2000) applied to Houfek, Rescigno & McCurdy’s two-dimensional resonant-collision model (2006). What appears to be new is that it is released as a library.

The closest relative: quantumGrid

quantumgrid — https://github.com/cwmccurdy/quantumGrid (fork: zstreeter/quantumGrid).

Authors	C. W. McCurdy, Z. Streeter, G. Barbalinardo (UC Davis / LBNL)
License	MIT
Latest release	2.0.30, April 2021 (PyPI); repository last updated December 2024
Stated purpose	“Exterior Complex Scaling Finite-Element Method Discrete Variable Representation grid for general physics problems” — written for a UC Davis graduate course in time-dependent quantum mechanics and released as open source

This is the same method lineage as qscat: McCurdy is a co-author of both the FEM-DVR-ECS method paper and the two-dimensional model qscat solves.

Shared with qscat (per its documentation and examples): Gauss–Lobatto FEM-DVR grid with bridge functions between elements; kinetic-energy matrix assembly in that basis; ECS applied as a complex final element; a Crank–Nicolson propagator; propagation on two coupled potential curves; a two-dimensional grid used for a two-electron problem (Hamiltonian_2D, its Two-electron_ECS example); built-in Morse and Coulomb potentials.

Not in it: any scattering observable. There is no driven-equation or Lippmann–Schwinger solve, no cross section of any kind, no local-complex-potential approximation, no resonance-pole search or complex quasi-bound levels, no propagator beyond Crank–Nicolson, no time-dependent energy-extraction transform, no sparse or out-of-core linear algebra, no grid-tuning, and no Coulomb (ionic) channels. Its two-dimensional example couples two electronic coordinates, not an electronic coordinate to a nuclear one, so it does not express the resonant electron–diatomic model at all.

Honest overlap: qscat.dvr and qscat.ecs re-implement numerics that quantumgrid already publishes. Both are independent implementations of the same published method, which makes quantumgrid a legitimate external cross-check for qscat’s one-dimensional grid — a use qscat does not currently make of it. Anyone who needs only a FEM-DVR-ECS grid for a one-dimensional problem, and especially anyone teaching the method, should look at quantumgrid first.

Other implementations of the same numerical ingredients

None of these compute electron–molecule scattering observables; they are grid, basis, or single-coordinate solvers.

Code	Language	What it is
COLOSS — Lei, Liu & Ren, Comput. Phys. Commun. (2025), arXiv:2407.16425	Fortran	Complex-scaled optical + Coulomb scattering solver for nuclear reactions; complex scaling to turn oscillatory boundary conditions into decaying ones
SEECS	Fortran	One-dimensional rovibrational Schrödinger equation with exterior complex scaling
FEDVR.jl	Julia	FE-DVR basis package (archived)
LBNL-AMO-MCTDHF	Fortran	MCTDHF electron dynamics from the same group; uses FEM-DVR internally, targets a different physical problem
Discvar, dvr_py, and ~15 similar repositories	Python	DVR basis/Hamiltonian construction; no ECS, no scattering
tRecX	C++	General time-dependent Schrödinger solver using infinite-range ECS for strong-field/attosecond problems

Electron–molecule scattering codes with a different method and scope

These solve the ab initio electron–molecule problem for real molecules — typically the fixed-nuclei electronic scattering problem, from which nuclear dynamics is treated separately or not at all. qscat solves a reduced, exactly-solvable model in which the electronic and nuclear coordinates are treated on the same footing, without the Born–Oppenheimer or local-complex-potential approximations. They are complements, not alternatives.

Code	Method	Availability
UKRmol+	Molecular R-matrix	Open source (Zenodo)
ePolyScat	Single-centre expansion / Schwinger variational	Open source; also hosted on the AMOS Gateway
FERM3D	Finite-element R-matrix	Published code
Quantemol-EC	Expert system driving UKRmol+	Commercial
PyOpenCAP	Complex absorbing potential on quantum-chemistry data	Open source; extracts resonance position and width, not cross sections
jitr	Calculable R-matrix on a Lagrange–Legendre mesh	Open source; nuclear physics

The resonance-model literature releases data, not solvers

The line of work qscat continues — the nonlocal resonance model and the exactly-solvable two-dimensional model (Domcke; Čížek, Horáček; Houfek, Rescigno & McCurdy) — has published extensively without releasing code. In particular no public repository, Zenodo deposit, or supplementary archive was found for:

• K. Houfek, T. N. Rescigno, C. W. McCurdy, Numerically solvable model for resonant collisions of electrons with diatomic molecules, Phys. Rev. A 73, 032721 (2006), https://doi.org/10.1103/PhysRevA.73.032721 — the model qscat solves;

• M. Váňa, K. Houfek, Phys. Rev. A 95, 022714 (2017), https://doi.org/10.1103/PhysRevA.95.022714 — the time-dependent formulation qscat’s TD route implements.

What this community distributes is cross-section data (LXCat, the Belgrade ACol/VAMDC database) rather than the solvers that produced it. The C++/CUDA code behind the second reference, from which qscat’s physics was re-derived, was never released either — qscat is its first public form.

What appears to have no released counterpart

Stated as survey results, not as priority claims:

• The two-dimensional resonant electron–diatomic model as a library. No released code was found that solves the electronic × nuclear driven problem and returns vibrational-excitation, dissociative-attachment, or dissociative-recombination cross sections.

• Exact solution and approximation side by side. qscat computes the exact two-dimensional result and the local-complex-potential approximation of the same quantity from one model definition, which is what makes the approximation testable rather than merely usable.

• Three time-dependent extractors driven by one propagation. The Tannor–Weeks, fixed-point (Dirac) and flux extractors share a single propagated wavepacket, so their disagreement measures discretisation error rather than differing dynamics. No released implementation of any of the three for this class of problem was found.

• An automatic FEM-DVR-ECS discretisation tuner. Grid parameters in this field are published as hand-tuned tables; no code was found that derives them from the potential and energy range.

• A complex-symmetric MUMPS backend with symbolic-analysis reuse across an energy sweep, exposed behind a solver-agnostic API with SuperLU as the fallback and differential oracle.

How this survey was done, and what it cannot see

Queries run on 2026-08-16: PyPI project and simple-index lookups; GitHub repository search for FEM-DVR, exterior complex scaling, discrete variable representation, electron molecule scattering, dissociative attachment, resonant electron scattering, vibrational excitation cross section; GitHub code search for "exterior complex scaling" and FEM-DVR; Zenodo software search; and literature search for released implementations of the nonlocal resonance model, the local-complex-potential model, and Tannor–Weeks cross-section extraction.

Notable negatives: the GitHub repository search returned no repositories whose name or description matches FEM-DVR, dissociative attachment, resonant electron scattering, or vibrational excitation cross section; the code search for "exterior complex scaling" returned essentially only qscat and quantumGrid. No Debian, Fedora, or conda-forge package exists for any code in this space.

Limits. This covers what is publicly indexed. It cannot see unreleased group codes (the Charles University and LBNL Fortran/C++ codes in this field are, as far as could be determined, unreleased), supplementary material behind paywalls, institutional repositories, or code shared privately on request. Treat “no released implementation found” as a statement about the public record on the survey date, not about what exists.

On the name

qscat here is quantum scattering, and the project’s domain is qscat.org — https://qscat.org is the project’s landing page and data.qscat.org serves the published artifacts. The acronym is used elsewhere for unrelated things — most visibly the QGIS Shoreline Change Analysis Tool (https://doi.org/10.1016/j.envsoft.2024.106263) and, historically, as shorthand for NASA’s QuikSCAT scatterometer mission. There is no functional overlap and no packaging collision: this project is repo-only and publishes nothing to PyPI, the PyPI distribution name qscat was unregistered as of 2026-08-16, and the QGIS tool ships through the QGIS plugin repository rather than PyPI. Readers searching literature or code for “QSCAT” should expect all three.