Public Beta
Fyron is in public beta. Research teams can use its documented workflows and examples, with the expectation that APIs and dependencies may change before version 1.0. Pin the version used for a study and review the release notes before upgrading.
Stability Expectations
| Area | Public-beta expectation |
|---|---|
| Core workflows | Tested and intended for real research workflows |
| API compatibility | Best effort within 0.x; breaking changes should be documented |
| Examples | Synthetic or de-identified only |
| Clinical deployment | Not supported |
| Regulated audit use | Not claimed |
Fyron favors small, explicit APIs and DataFrame-friendly outputs. When an API changes, the project should prefer additive options, compatibility aliases, or release notes that explain the migration.
Responsible Use
Fyron can help structure clinical data science work, but users remain responsible for:
- study approvals and institutional governance,
- privacy review and de-identification,
- clinical validation and endpoint definitions,
- correct interpretation of statistical and ML outputs,
- secure handling of FHIR, DICOMweb, S3, and other credentials.
Do not share PHI, secrets, private endpoint URLs, or real patient-linked DICOM/FHIR payloads in public issues, pull requests, screenshots, or examples.
Documentation and Support
- Getting Started covers installation and a first workflow.
- Research Reliability records verification evidence, platform coverage, and known boundaries.
- Release History documents changes, corrections, and upgrade requirements.
- The support guide explains how to ask questions and report reproducible issues.
Current Limitations
- API documentation is still being expanded; some generated function descriptions are incomplete.
- Optional extras have dependency and platform requirements. Check Installation & Extras for the workflow you intend to use.
- Configured CI jobs do not establish support for every platform. See the dated verification evidence in Research Reliability.
When upgrading, retain the original data, parameters, and environment information needed to reproduce your analysis. Release notes identify known corrections that may require earlier results to be recalculated.