The hardware and bandwidth for this mirror is donated by dogado GmbH, the Webhosting and Full Service-Cloud Provider. Check out our Wordpress Tutorial.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]dogado.de.

fdars 0.3.2

Bug Fixes

• Fixed compiled code WARNING: wrapped abort/exit/_exit symbols using linker --wrap flag to convert process termination into R errors
• Reduced test CPU time by limiting Rust thread pool (RAYON_NUM_THREADS=2)

fdars 0.3.1

Bug Fixes

• Fixed Windows installation failure (missing cargo checksum files)
• Wrapped slow bootstrap examples in \donttest{}

fdars 0.3.0

Internal

• Upgraded Rust backend (fdars-core) to v0.4.0
• New FdMatrix type for safer matrix handling (internal)
• New streaming depth module in core (internal)
• Reduced package size by removing non-essential vendored files
• No user-facing API changes

fdars 0.2.0

Test Coverage & Quality

• Improved Rust core test coverage to 84%+
• Improved R package test coverage to 80%+
• Added pre-commit hooks for cargo fmt and clippy

New Features

Optimal Cluster Selection

• Added optim.kmeans.fd() function to automatically determine the optimal number of clusters for functional k-means
• Three selection criteria available:
  • Silhouette score: Measures cluster cohesion vs separation (-1 to 1, higher is better)
  • Calinski-Harabasz index: Ratio of between/within cluster variance (higher is better)
  • Elbow method: Visual inspection of within-cluster sum of squares
• Added print() and plot() methods for optim.kmeans.fd objects
• Silhouette and Calinski-Harabasz computations implemented in Rust for performance

k-NN Bandwidth Selection for Nonparametric Regression

• Added k-nearest neighbors support to fregre.np() via the type.S parameter:
  • "kNN.gCV": Global cross-validation (single k for all observations)
  • "kNN.lCV": Local cross-validation (adaptive k per observation)
• Extended predict.fregre.np() to handle k-NN models

Flexible Metrics in Clustering

• kmeans.fd() now accepts both string metrics ("L2", "L1", "Linf") and metric/semimetric functions
• String metrics use fast Rust-only path; function metrics provide flexibility for custom distances

Improvements

ggplot2 Visualizations

• All plot methods now use ggplot2 instead of base R graphics:
  • plot.fdata(): Functional data curves with minimal theme
  • plot.kmeans.fd(): Cluster-colored curves with dashed cluster centers
  • plot.optim.kmeans.fd(): Criterion scores with optimal k highlighted
  • plot.outliers.fdata(): Outlier/normal curves with color legend

Documentation

Vignettes

Added 6 comprehensive vignettes: - Introduction to fdars - Functional Depth Functions - Distance Metrics and Semimetrics - Functional Regression - Functional Clustering - Outlier Detection

API Documentation

• Complete roxygen2 documentation for all exported functions

Bug Fixes

• Fixed namespace issues with stats and utils imports
• Fixed ggplot2 .data pronoun import for R CMD check compliance

fdars 0.1.0

• Initial release
• Core functional data class (fdata) with 1D and 2D support
• 7 depth functions: FM, mode, RP, RT, FSD, KFSD, RPD
• Distance metrics: Lp, Hausdorff, DTW, KL
• Semimetrics: PCA, derivative, basis, Fourier, hshift
• Functional regression: PC, basis, nonparametric
• K-means clustering with k-means++ initialization
• Outlier detection: depth-based and LRT methods
• Statistical tests: flm.test, fmean.test
• Bootstrap inference and confidence intervals
• High-performance Rust backend with parallel processing

These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.
Health stats visible at Monitor.