Package index • songR

Core Functions

Fit a SONG model, incrementally update it with new data, and project unseen points into the learned embedding space.

song(): Fit a SONG Model for Dimensionality Reduction
update(<song_model>): Incrementally Update a SONG Model with New Data
predict(<song_model>): Project New Points into a SONG Embedding

Visualization

Plot SONG embeddings, codebook prototypes, and the learned topology graph using base R or ggplot2.

plot(<song_model>): Plot a SONG Model
autoplot.song_model(): Autoplot Method for SONG Models

Model Inspection

Print and summarize SONG model objects, including codebook statistics, quantization error, and graph connectivity.

print(<song_model>): Print a SONG Model
summary(<song_model>): Summarize a SONG Model

Interactive App

Launch a Shiny app to compare SONG, t-SNE, and UMAP side-by-side with dark mode, custom data upload, and incremental update support.

run_songR_app(): Launch Interactive Comparison App

Data

Bundled datasets for examples, vignettes, and benchmarking.

songR_blobs: Simulated Gaussian Blobs Dataset

Package

songR songR-package: songR: Self-Organizing Nebulous Growths for Dimensionality Reduction