GPU-LMFit

Lightweight Levenberg–Marquardt fitting library with CUDA-ready GPU codepaths.

This repository contains a small fitting library (CPU + GPU sources) and an simple example CPU program that fits a 1D Gaussian.

Current status

• CPU example: working (examples/CPU/Gauss_1D)
• GPU fitting: in progress — GPU kernels and CUDA sources are present in src/gpu and will be wired up soon :)

Build (Windows, out-of-source)

This project uses CMake. The example was developed using Ninja as the generator but any generator supported by your platform should work.

From the repository root (PowerShell):

mkdir build; cd build
cmake -G Ninja ..
cmake --build .

If you don't have Ninja, replace -G Ninja with the generator you prefer (like "Visual Studio 16 2019").

Run the CPU example

After building, the CPU example executable Gauss_1D is available in the build subdirectory examples/CPU. It uses the headers from include/ and the CPU solver implementation from src/cpu.

From the build directory you can run (adjust path if your generator places binaries elsewhere):

.\examples\CPU\Gauss_1D.exe

Project layout

• include/ — public headers (models, types, utilities)
• src/ — implementation (cpu, gpu, binding)
  • cpu/ — CPU solver and example wiring
  • gpu/ — CUDA kernels and GPU estimators (work in progress)
• examples/ — small example programs (currently CPU Gauss_1D)
• tests/ — test sources (experimental)

Roadmap / Notes / WIP

• Finish GPU integration and provide a GPU example binary.
• (Add unit tests and CI for cross-platform builds.)

License

See the LICENSE file in the repository root.