iints-sdk-python35 1.0.0

A pre-clinical Edge-AI SDK for diabetes management validation.

IINTS-AF SDK

IINTS-AF is a safety-first SDK for insulin-algorithm research. It lets you simulate, validate, and report results with reproducible artifacts.

Docs (GitHub Pages): python35.github.io/IINTS-SDK

What You Can Do

• Run virtual patient simulations.
• Test algorithm safety gates (deterministic supervisor).
• Add optional AI glucose forecasting.
• Validate datasets before training/evaluation.
• Generate audit-ready CSV/JSON/PDF/HTML outputs.

Quick Start

python3 -m venv .venv
source .venv/bin/activate
python -m pip install -U pip
pip install iints-sdk-python35

iints doctor --smoke-run
iints quickstart --project-name iints_quickstart
cd iints_quickstart
iints presets run --name baseline_t1d --algo algorithms/example_algorithm.py

MDMP (Short)

MDMP is the data-quality protocol used by IINTS.

• Contract: defines expected columns, types, units, and bounds.
• Validation: checks a dataset against the contract.
• Fingerprint + Grade: writes deterministic hashes and a grade (draft, research_grade, clinical_grade).
• Visualizer: builds a single-file HTML report for audits.

Use the dedicated namespace:

iints mdmp template --output-path mdmp_contract.yaml
iints mdmp validate mdmp_contract.yaml data/my_cgm.csv --output-json results/mdmp_report.json
iints mdmp visualizer results/mdmp_report.json --output-html results/mdmp_dashboard.html

Use standalone MDMP backend (optional):

export IINTS_MDMP_BACKEND=mdmp_core

Staleness / lineage checks (standalone MDMP CLI):

mdmp fingerprint-record data/my_cgm.csv --output-json results/fingerprint.json --expires-days 365
mdmp fingerprint-check results/fingerprint.json data/my_cgm.csv
mdmp lineage-card-refresh results/mdmp_model_card.yaml
mdmp registry init --registry registry/mdmp_registry.json
mdmp registry push --registry registry/mdmp_registry.json --report results/mdmp_report.json

Dual Repo Workflow

• SDK repo: python35/IINTS-SDK
• MDMP repo: python35/MDMP

Local helper scripts:

• tools/local/dual_repo_status.sh
• tools/local/dual_repo_commit_push.sh

Full process: docs/DUAL_REPO_WORKFLOW.md

MDMP sync CI gate:

• .github/workflows/mdmp-sync.yml
• Pulls the public python35/MDMP repo directly.
• Auto dependency updates for MDMP are handled via Dependabot (.github/dependabot.yml).

Typical Workflow

1. Prepare or import data.
2. Validate data with MDMP.
3. Run simulation or forecast evaluation.
4. Review report artifacts and metrics.

Key Commands

iints run-full --algo algorithms/example_algorithm.py --scenario-path scenarios/clinic_safe_baseline.json --output-dir results/run_full
iints scorecard --algo algorithms/example_algorithm.py --profile research_default --output-dir results/scorecard
iints study-ready --algo algorithms/example_algorithm.py --output-dir results/study_ready
iints sources --output-json results/source_manifest.json

Documentation

• Docs site: python35.github.io/IINTS-SDK
• Plain guide: docs/PLAIN_LANGUAGE_GUIDE.md
• Comprehensive guide: docs/COMPREHENSIVE_GUIDE.md
• MDMP draft: docs/MDMP.md
• Demos: examples/demos
• Notebooks: examples/notebooks

Safety Notice

For research use only. Not a medical device. No clinical dosing advice.