|
fastlowess-cpp
High-performance LOWESS smoothing
|
Run the batch LOWESS fit on the GPU via wgpu (Vulkan/Metal/DX12, GLES on Android) instead of the CPU.
The batch Lowess type can execute on a GPU-accelerated backend powered by wgpu. It reimplements almost the entire LOWESS pipeline — local regression fitting, robustness iterations, interval bounds, and cross-validation — as WGSL compute shaders, so all anchor points are fit in parallel instead of one at a time on CPU cores.
This is worth enabling for high-throughput processing of large datasets (roughly 10k+ points); for smaller inputs the CPU backend (optionally with parallel = true) is typically faster once you account for GPU dispatch overhead. See BENCHMARKS.md for crossover points measured on real hardware.
Batch only. GPU support applies to the batch
Lowesstype only.StreamingLowess/OnlineLowessremain CPU-only — the Rust core documents GPU as optimized for static batch data, not incremental chunk/point processing.
GPU support is opt-in and not included in prebuilt C++ releases — download a prebuilt GPU-enabled build via the one-time installer below, or build from source with the gpu Cargo feature.
tricube, epanechnikov, gaussian, uniform, biweight, triangle, cosine).bisquare, huber, and talwar robustness weighting.mad (Median Absolute Deviation), mar (Median Absolute Residual), and mean (Mean Absolute Residual).delta parameter to accelerate smoothing on dense grids.kfold and loocv cross-validation.| Feature | CPU | GPU | Notes |
|---|---|---|---|
| Batch fitting | ✅ | ✅ | |
| Streaming/Online | ✅ | ❌ | GPU optimized for static batch data |
| All weight/robustness/scaling methods | ✅ | ✅ | |
| Confidence/prediction intervals | ✅ | ✅ | |
| Cross-validation (k-fold, LOOCV) | ✅ | ✅ | |
| Unsorted inputs and sorted-output option | ✅ | ✅ | GPU results preserve Batch output ordering. |
| Custom per-observation weights | ✅ | ✅ | Applied in GPU fit and CV candidate kernels. |
Each compute pipeline now binds only its own resources, so adapter limits are based on the largest individual GPU pass rather than the union of all shader bindings.
Before requesting backend = "gpu", check whether the currently loaded library was built with GPU support. Requesting the GPU backend when it isn't available raises a clear error pointing at fastlowess::gpu::install(), rather than a raw panic.
This binding ships a one-time installer that downloads a prebuilt GPU-enabled build from the gpu-builds release (built by .github/workflows/release-gpu.yml) — a single perpetual release holding GPU artifacts for every version, so individual version release pages stay uncluttered; the source version is embedded in each asset's filename instead. Building from source with the gpu Cargo feature is always available as an alternative.
Non-interactively: fastlowess::gpu::install(/*yes=*/true). Requires curl on PATH (ships with Linux, macOS, and Windows 10+). A running process cannot swap the backend of a library it already linked against — after downloading, relink/rebuild your application against the downloaded library (or dlopen/LoadLibrary it manually) and restart.
Build from source instead:
Once GPU support is available, request it by setting the backend option on the batch constructor.
If GPU support isn't available, requesting backend = "gpu" raises a runtime error pointing at fastlowess::gpu::install() rather than a raw Rust panic.
The GPU backend leverages wgpu and supports:
It requires a device supporting compute shaders. If no compatible GPU is found at runtime, model construction raises an error.
The GPU backend is optimized for large datasets (N > 100,000) and provides parallelization through compute shaders. For smaller datasets, the CPU backend (backend = "cpu", the default) is faster.