MLflow R Starter

A Docker-based starter for MLflow experiment tracking with R. Uses the iris dataset and random forest as a demonstration - replace with your own data and models.

Quick Start

docker-compose up

Open http://localhost:5001 to view MLflow UI.

Structure

• R/mlflow_example.R - Main R script with MLflow integration
• docker-compose.yml - MLflow server + R service
• docker/r-runtime/Dockerfile - R environment

Customize

Replace the Iris dataset and Random Forest model in mlflow_example.R with your data and algorithm. Update the TODO comments throughout the file.

Notes

There are some limitations to logging models with R around model signatures: mlflow/mlflow#4462