Roadmap¶

The priority remains reliable scientific interpretation, not the number of experimental APIs.

Delivered to users¶

Published stable 1.2.0 (24 October 2026): native continuous/sparse methods, qualified 1.1 module additions, ten explicit W1–W4 orchestration routes, data validation and documentation for Python 3.11–3.13.

Install 1.2.0 · Release notes · Choose a workflow

The frozen package-root boundary is unchanged, and the separate E1–E4 helpers remain experimental.

Research and documentation delivered¶

W3 R76–R95 synthetic simulation, falsification, missingness and identifiability records remain in the study map and figure atlas. Publication does not convert exploratory scientific evidence into validated coverage claims.

Open research priorities — not blockers to stable 1.2.0¶

• Establish reliable simultaneous functional confidence bands under extreme leverage, with independent-source evaluation: #289.
• Clarify externally defensible audit/missingness mechanisms before proposing outcome-dependent gaze corrections: #296.
• Qualify any future E1–E4 promotion separately.
• Require a fresh exact-version release process for any future Python distribution.

Stable 1.2.0 publication is complete. This post-publication closeout changes no scientific estimators.