LIBSkit

A software tool for the simulation and analysis of laser-induced breakdown spectroscopy (LIBS) spectra.

The simulation core is written in C++ with Python bindings via pybind11. Atomic databases (NIST ASD lines, ionization energies, energy levels) are bundled with the package and loaded automatically on import.

Quick Start

import libskit
import numpy as np

# Create an LTE plasma model
model = libskit.LTEModel(
    {"Si": 0.72, "Na": 0.09, "Ca": 0.19},   # %
    temperature=17300.0,                    # K
    electron_density=2.53e23,               # m^-3
    length=2.4e-3                           # m
)

# Define an instrument and calculate a spectrum
wavelengths = np.arange(250, 900, 0.05)                       # nm
instrument = libskit.GenericSpectrometer(0.05, 0.001)         # nm
spectrum = instrument.calculate_spectrum(wavelengths, model)

Installation

This package is only available as a source build — it is not published on PyPI or any package registry.

Source builds need a Clang toolchain with the Enzyme automatic-differentiation plugin.

• Linux: sudo apt install cmake ninja-build git clang llvm-dev
• macOS: install LLVM/Clang and CMake, then build Enzyme against that LLVM
• Windows: use an LLVM/Clang toolchain with Enzyme available

Then:

git clone <repository-url>
cd libskit

python -m venv .venv
source .venv/bin/activate  # Linux/macOS
# .venv\Scripts\activate   # Windows

pip install -r requirements.txt
./scripts/dev.sh

This builds the Enzyme plugin and installs the package in one step.

Note: Eigen3 is downloaded automatically during build if not found on your system.

Prerequisites (for development)

To build the code from source you need:

• Clang-compatible C++20 compiler
• CMake 3.15+
• Python 3.8+

Tutorials

The notebooks/ directory in this repository contains a curated set of tutorials (tutorial_1_… through tutorial_7_…).

Contributors

Christoph H. Egerland

License

MIT