R Package

Rhobots

The full BERTopic pipeline — transformer embeddings, UMAP, HDBSCAN density clustering, and c-TF-IDF topic representation — running entirely in R. No Python. No conda. No reticulate.

CRAN v0.1.6 R ≥ 4.1 MIT License Windows · macOS · Linux
📝
Embed
torch · safetensors
📉
Reduce
uwot · UMAP
🔵
Cluster
Rcpp · HDBSCAN
🏷️
Represent
c-TF-IDF · POS
Resources
Hands-on

Full Tutorial

End-to-end walkthrough on 3,328 African research abstracts — embeddings, parameter sweep, topic fitting, and interactive visualisations.

How it works

How UMAP Works

Illustrated technical explainer: manifold hypothesis, fuzzy membership weights, SGD optimisation, and the effect of every parameter.

How it works

How HDBSCAN Works

Core distances, mutual reachability, Borůvka MST, the condensed cluster tree, EOM extraction, and what every parameter controls.

1
Install from CRAN
install.packages("Rhobots")
2
Install the torch backend (once per machine, ~560 MB)
library(Rhobots)
rhobots_install()   # restart R when prompted
3
Try the built-in demo
rhobots_demo()   # downloads Gutenberg books, fits a model, opens a plot
Citation
van der Pol, J.P.G. (2026). Rhobots: BERTopic-Style Topic Modeling Without Python.
R package version 0.1.6. https://github.com/JPvdP/Rhobots