autoSpectra v0.3.0 — CI/CD, full documentation & Zenodo

autoSpectra v0.3.0

What's new in v0.3.0

• ✅ R-CMD-check passing on all 4 runners: ubuntu, windows, macOS (R release) + ubuntu (R-devel)
• 📚 44 functions fully documented via roxygen2 (all .Rd files committed)
• ⚙️ GitHub Actions CI/CD: R-CMD-check.yml + pkgdown.yml (auto-deploy on every push)
• 🌐 pkgdown website live at https://HugoMachadoRodrigues.github.io/autoSpectra/
• 🔧 Fixed all R CMD check WARNINGs (portable filenames, import declarations, Unicode escapes in Rd)
• 📦 Added .Rbuildignore, DCF-format LICENSE, globalVariables() for ggplot2 NSE
• 🗄️ Fixed .zenodo.json relation types for valid Zenodo metadata

Core features (inherited from v0.2.0)

• OSSL v1.2 integration — sensor-agnostic soilVAE models for VisNIR (350–2500 nm) and MIR (600–4000 cm⁻¹)
• 34 OSSL Level-1 harmonized soil properties (organic carbon, texture, pH, CEC, micronutrients, and more)
• Two-step Savitzky-Golay pipeline: SG smooth → SG 1st derivative
• soilVAE asymmetric autoencoder with conformal prediction intervals and Mahalanobis applicability domain
• Shiny interface for interactive prediction and visualization

Data source

Trained on the Open Soil Spectral Library (OSSL) v1.2
Safanelli et al. (2023), Earth System Science Data, doi:10.5194/essd-15-3829-2023

Installation

devtools::install_github("HugoMachadoRodrigues/autoSpectra")

Citation

Rodrigues, H. (2026). autoSpectra: Soil Spectral Modelling, Visualization and Prediction (v0.3.0). Zenodo.

@software{Rodrigues2026autoSpectra,
  author  = {Rodrigues, Hugo},
  orcid   = {0000-0002-8070-8126},
  title   = {autoSpectra: Soil Spectral Modelling, Visualization and Prediction},
  year    = {2026},
  version = {0.3.0},
  url     = {https://github.com/HugoMachadoRodrigues/autoSpectra}
}

autoSpectra v0.2.0 — R package + OSSL v1.2 integration

autoSpectra v0.2.0

First stable release of the autoSpectra R package for soil spectral modelling, visualization and prediction.

Highlights

• OSSL v1.2 integration — sensor-agnostic soilVAE models for VisNIR (350–2500 nm) and MIR (600–4000 cm⁻¹)
• 34 OSSL Level-1 harmonized soil properties (organic carbon, texture, pH, CEC, micronutrients, and more)
• Two-step Savitzky-Golay pipeline: SG smooth → SG 1st derivative
• soilVAE asymmetric autoencoder with conformal prediction intervals and Mahalanobis applicability domain
• Shiny interface for interactive prediction and visualization
• Full roxygen2 documentation (44 exported functions), R-CMD-check clean on ubuntu/windows/macOS + R-devel

Data source

Trained on the Open Soil Spectral Library (OSSL) v1.2
Safanelli et al. (2023), Earth System Science Data, doi:10.5194/essd-15-3829-2023

Installation

devtools::install_github("HugoMachadoRodrigues/autoSpectra")

Citation

If you use this software, please cite:

Rodrigues, H. (2026). autoSpectra: Soil Spectral Modelling, Visualization and Prediction (v0.2.0). Zenodo. https://doi.org/10.5281/zenodo.XXXXXXX

@software{Rodrigues2026autoSpectra,
  author  = {Rodrigues, Hugo},
  orcid   = {0000-0002-8070-8126},
  title   = {autoSpectra: Soil Spectral Modelling, Visualization and Prediction},
  year    = {2026},
  version = {0.2.0},
  doi     = {10.5281/zenodo.XXXXXXX},
  url     = {https://github.com/HugoMachadoRodrigues/autoSpectra}
}