Roadmap

The roadmap outlines upcoming improvements. Priorities may change as new feedback arrives.

In progress

• Validated branch aggregation: research same-class aggregation that recomputes distributions and meets compression, quality, stability, and runtime gates.
• Improved Spark support: streamline the interface for PySpark DataFrames and add automated schema validation.
• Interactive dashboards: publish Streamlit templates to explore clusters and export rule summaries.

Planned

• Model explainers: integrate SHAP-style visualizations tailored to InsideForest regions.
• Time-series helpers: add utilities to segment temporal data with sliding windows.
• Rust backend prototype: experiment with a compiled implementation to reduce inference latency.

Community requests

• Share ideas or vote on priorities via GitHub issues.
• Tag proposals with roadmap and describe the use case, dataset size, and desired outcome.

Recently delivered

• Canonical supervised region clusterers with explicit unmatched assignments and no forest fallback.
• Language toggle across the documentation site.
• Expanded API reference with cross-links between modules.
• New reproducibility guidelines covering seeding, logging, and automation.