nirs4all is an open-source ecosystem for NIR spectroscopy workflows: use Studio, share UI components, consume the portable core aggregate, read vendor files, assemble datasets, run chemometric methods, publish reproducible pipelines, benchmark results, and inspect ecosystem health.
Use this page when searching for open-source NIRS software, Studio tooling, the portable core aggregate, shared UI components, provider clients, near-infrared spectroscopy file readers, PLS tools, reproducible NIRS datasets, or browser-based spectroscopy modelling.
nirs4allPython reference library for preprocessing spectra, building DAG-ML pipelines, visualizing results, and deploying NIRS models. An R package and WebAssembly bindings share its native method engine, so recipes and trained models move between Python, R and the browser.
Open nirs4all.org
nirs4all-studioExplore spectra, build pipelines and inspect predictions without writing code. Studio 0.14.0 runs on a native backend with a bundled nirs4all 1.3 runtime and offers every nirs4all-methods estimator as a pipeline node. One installer per operating system and a Docker image are available.
Studio 0.14.0 installers and Docker
quali-nirs4allBeta guided browser application for NIRS labs: check data health and repeats, select HPLC or wet-chemistry samples, calibrate, qualify predictions, and maintain models through a pedagogical workflow.
Open quali.nirs4all.org Open the source repository
nirs4all-devicePrototype phone and tablet workbench for connected NIRS spectrometers, with Bluetooth capture, field sessions, quality gates, local storage, and portable inference. Hardware files and a build guide for a DIY NIRS device are coming soon.
Open device.nirs4all.org Open the source repository
nirs4all-corePublished portable V1 aggregate over dag-ml, dag-ml-data, formats, IO, datasets and methods, providing one coherent cross-language integration surface.
Open nirs4all-core releases Open the source repository
nirs4all-uiPublished React/TypeScript package of reusable components, runtime status helpers, score helpers, default CSS tokens, brand generators, and packaged visual assets for Studio, Web, and custom app hosts.
Open component showcase
nirs4all-providersDependency-light Python provider clients over datasets and repository contracts. Benchmarks and papers keep their own public APIs. Read-side only, soft-importing, with package and source-repository surfaces.
Open provider page Open source repo
nirs4all-toolsPublished V1 preservation toolkit for read-only inspection and checksummed copy-only migration with dry-run, resume and verify. Semantic workspace conversion fails closed instead of claiming unsupported completeness.
Open nirs4all-tools releases Open nirs4all-tools on PyPI
nirs4all-formatsRust and WebAssembly readers for vendor NIRS and spectroscopy file formats, with an in-browser demo that keeps files local.
Open formats.nirs4all.org
nirs4all-datasetsCitable catalog of raw NIRS reference datasets with provenance, access tiers, statistics, DOI links, and byte-free identity cards.
Open datasets.nirs4all.org
nirs4all-methodsPublished portable C++17 libn4m methods engine with a stable C ABI, language bindings, parity reports, and scientific documentation mapping 209 method pages to 88 bibliography entries.
Open methods.nirs4all.org nirs4all-aomPython/sklearn package for AOM-PLS, POP-PLS, AOM-Ridge and FastAOM models used by the adaptive-model paper companion workflows.
Open nirs4all-aom on PyPI Open source repo
nirs4all-webFull-WASM browser workbench for uploading spectra, building a pipeline, running cross-validated PLS or PLS-DA, and predicting locally.
Open web.nirs4all.org
nirs4all-clusterTrusted-LAN beta for distributing nirs4all execution across lab machines with a FastAPI coordinator, SQLite queue, capability routing, and crash-safe leases. It is a validation bench, not a public multi-tenant service.
Open nirs4all-cluster
nirs4all-ioDataset assembly bridge and browser dataset builder for turning spectra and sidecars into validated nirs4all dataset specifications.
Open io.nirs4all.org
nirs4all-repositoryPublic beta catalog of pre-configured NIRS pipeline recipes for nirs4all and dag-ml, with provenance, checksums, and per-recipe validation status.
Open repository.nirs4all.org
nirs4all-papersReproduction-document publisher for deposited papers, methods, bibliography, sidecars, and live in-browser pipeline replay.
Open papers.nirs4all.org
nirs4all-benchmarksPrototype static benchmark arena for scored, reproducible NIRS pipeline comparisons and dataviz.
Open benchmarks.nirs4all.org nirs4all-cockpitOperational dashboard for registries, releases, code statistics, CI, documentation, and public page visits across the ecosystem.
Open cockpit.nirs4all.org Open the initial V1 release receipt| Task | Use | Tool |
|---|---|---|
| Read vendor spectra | Decode spectroscopy files locally in the browser or from Rust bindings. | nirs4all-formats |
| Build datasets | Combine spectra, targets, metadata, sidecars, and validations into one dataset specification. | nirs4all-io |
| Find reference data | Browse NIRS datasets with provenance, license, DOI, and statistical identity cards. | nirs4all-datasets |
| Run methods | Use portable PLS and NIRS methods through C, Python, R, MATLAB, JavaScript, and other bindings. | nirs4all-methods |
| Work in R | Run NIRS pipelines on the same native methods from R, and exchange recipes and trained models with Python and the browser. | nirs4all for R |
| Run adaptive models | Use the paper companion implementation of AOM-PLS, POP-PLS, AOM-Ridge and FastAOM estimators. | nirs4all-aom |
| Use a desktop app | Build and inspect NIRS workflows with Studio 0.14.0. | Studio 0.14.0 installers and Docker |
| Guide routine lab work | Assess spectral data, select reference analyses, calibrate, qualify predictions, and maintain NIRS models in a guided browser workflow. | quali-nirs4all (beta) |
| Capture spectra in the field | Connect a mobile device to a Bluetooth NIRS spectrometer, organize sessions, check capture quality, and run portable inference. | nirs4all-device (prototype) |
| Use the portable aggregate | Integrate the published DAG-ML, data, format, IO, dataset, and method stack through one package. | nirs4all-core |
| Share UI components | Reuse the published Studio/Web package, runtime badges, status view models, score helpers and brand assets. | nirs4all-ui |
| Read provider contracts | Consume datasets and repository descriptors through one soft-importing read-side client layer; benchmarks and papers keep their own APIs. | nirs4all-providers |
| Preserve saved artifacts | Inspect legacy workspaces, then create a disjoint, checksummed copy-only archive. Semantic conversion is explicitly unavailable in V1. | nirs4all-tools |
| Model in the browser | Upload spectra, build a pipeline, run cross-validation, inspect results, and predict without a backend. | nirs4all-web |
| Distribute lab execution | Route trusted-LAN jobs across local lab machines with scheduler, leases, capability matching, and crash recovery. | nirs4all-cluster |
| Reuse pipelines | Browse pipeline recipes with provenance and checksums; validation status is recorded per recipe. | nirs4all-repository |
| Publish reproducible work | Deposit paper bundles with methods, bibliography, reproducibility sidecars, and live replay. | nirs4all-papers |
| Monitor quality | Track benchmarks, package health, releases, CI, and ecosystem status. | benchmarks and cockpit |