Metadata-Version: 2.4
Name: vahtian-epinet
Version: 0.4.3
Summary: EpiNet (Epistemic Network): transparent network/feature-space analysis with honest evaluation (calibration, permutation nulls, model cards), nearest-centroid contestability, and federated governed aggregates.
Author: EpiNet contributors
License:                                  Apache License
                                   Version 2.0, January 2004
                                http://www.apache.org/licenses/
        
           TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
        
           1. Definitions.
        
              "License" shall mean the terms and conditions for use, reproduction,
              and distribution as defined by Sections 1 through 9 of this document.
        
              "Licensor" shall mean the copyright owner or entity authorized by
              the copyright owner that is granting the License.
        
              "Legal Entity" shall mean the union of the acting entity and all
              other entities that control, are controlled by, or are under common
              control with that entity. For the purposes of this definition,
              "control" means (i) the power, direct or indirect, to cause the
              direction or management of such entity, whether by contract or
              otherwise, or (ii) ownership of fifty percent (50%) or more of the
              outstanding shares, or (iii) beneficial ownership of such entity.
        
              "You" (or "Your") shall mean an individual or Legal Entity
              exercising permissions granted by this License.
        
              "Source" form shall mean the preferred form for making modifications,
              including but not limited to software source code, documentation
              source, and configuration files.
        
              "Object" form shall mean any form resulting from mechanical
              transformation or translation of a Source form, including but
              not limited to compiled object code, generated documentation,
              and conversions to other media types.
        
              "Work" shall mean the work of authorship, whether in Source or
              Object form, made available under the License, as indicated by a
              copyright notice that is included in or attached to the work
              (an example is provided in the Appendix below).
        
              "Derivative Works" shall mean any work, whether in Source or Object
              form, that is based on (or derived from) the Work and for which the
              editorial revisions, annotations, elaborations, or other modifications
              represent, as a whole, an original work of authorship. For the purposes
              of this License, Derivative Works shall not include works that remain
              separable from, or merely link (or bind by name) to the interfaces of,
              the Work and Derivative Works thereof.
        
              "Contribution" shall mean any work of authorship, including
              the original version of the Work and any modifications or additions
              to that Work or Derivative Works thereof, that is intentionally
              submitted to Licensor for inclusion in the Work by the copyright owner
              or by an individual or Legal Entity authorized to submit on behalf of
              the copyright owner. For the purposes of this definition, "submitted"
              means any form of electronic, verbal, or written communication sent
              to the Licensor or its representatives, including but not limited to
              communication on electronic mailing lists, source code control systems,
              and issue tracking systems that are managed by, or on behalf of, the
              Licensor for the purpose of discussing and improving the Work, but
              excluding communication that is conspicuously marked or otherwise
              designated in writing by the copyright owner as "Not a Contribution."
        
              "Contributor" shall mean Licensor and any individual or Legal Entity
              on behalf of whom a Contribution has been received by Licensor and
              subsequently incorporated within the Work.
        
           2. Grant of Copyright License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              copyright license to reproduce, prepare Derivative Works of,
              publicly display, publicly perform, sublicense, and distribute the
              Work and such Derivative Works in Source or Object form.
        
           3. Grant of Patent License. Subject to the terms and conditions of
              this License, each Contributor hereby grants to You a perpetual,
              worldwide, non-exclusive, no-charge, royalty-free, irrevocable
              (except as stated in this section) patent license to make, have made,
              use, offer to sell, sell, import, and otherwise transfer the Work,
              where such license applies only to those patent claims licensable
              by such Contributor that are necessarily infringed by their
              Contribution(s) alone or by combination of their Contribution(s)
              with the Work to which such Contribution(s) was submitted. If You
              institute patent litigation against any entity (including a
              cross-claim or counterclaim in a lawsuit) alleging that the Work
              or a Contribution incorporated within the Work constitutes direct
              or contributory patent infringement, then any patent licenses
              granted to You under this License for that Work shall terminate
              as of the date such litigation is filed.
        
           4. Redistribution. You may reproduce and distribute copies of the
              Work or Derivative Works thereof in any medium, with or without
              modifications, and in Source or Object form, provided that You
              meet the following conditions:
        
              (a) You must give any other recipients of the Work or
                  Derivative Works a copy of this License; and
        
              (b) You must cause any modified files to carry prominent notices
                  stating that You changed the files; and
        
              (c) You must retain, in the Source form of any Derivative Works
                  that You distribute, all copyright, patent, trademark, and
                  attribution notices from the Source form of the Work,
                  excluding those notices that do not pertain to any part of
                  the Derivative Works; and
        
              (d) If the Work includes a "NOTICE" text file as part of its
                  distribution, then any Derivative Works that You distribute must
                  include a readable copy of the attribution notices contained
                  within such NOTICE file, excluding those notices that do not
                  pertain to any part of the Derivative Works, in at least one
                  of the following places: within a NOTICE text file distributed
                  as part of the Derivative Works; within the Source form or
                  documentation, if provided along with the Derivative Works; or,
                  within a display generated by the Derivative Works, if and
                  wherever such third-party notices normally appear. The contents
                  of the NOTICE file are for informational purposes only and
                  do not modify the License. You may add Your own attribution
                  notices within Derivative Works that You distribute, alongside
                  or as an addendum to the NOTICE text from the Work, provided
                  that such additional attribution notices cannot be construed
                  as modifying the License.
        
              You may add Your own copyright statement to Your modifications and
              may provide additional or different license terms and conditions
              for use, reproduction, or distribution of Your modifications, or
              for any such Derivative Works as a whole, provided Your use,
              reproduction, and distribution of the Work otherwise complies with
              the conditions stated in this License.
        
           5. Submission of Contributions. Unless You explicitly state otherwise,
              any Contribution intentionally submitted for inclusion in the Work
              by You to the Licensor shall be under the terms and conditions of
              this License, without any additional terms or conditions.
              Notwithstanding the above, nothing herein shall supersede or modify
              the terms of any separate license agreement you may have executed
              with Licensor regarding such Contributions.
        
           6. Trademarks. This License does not grant permission to use the trade
              names, trademarks, service marks, or product names of the Licensor,
              except as required for reasonable and customary use in describing the
              origin of the Work and reproducing the content of the NOTICE file.
        
           7. Disclaimer of Warranty. Unless required by applicable law or
              agreed to in writing, Licensor provides the Work (and each
              Contributor provides its Contributions) on an "AS IS" BASIS,
              WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
              implied, including, without limitation, any warranties or conditions
              of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
              PARTICULAR PURPOSE. You are solely responsible for determining the
              appropriateness of using or redistributing the Work and assume any
              risks associated with Your exercise of permissions under this License.
        
           8. Limitation of Liability. In no event and under no legal theory,
              whether in tort (including negligence), contract, or otherwise,
              unless required by applicable law (such as deliberate and grossly
              negligent acts) or agreed to in writing, shall any Contributor be
              liable to You for damages, including any direct, indirect, special,
              incidental, or consequential damages of any character arising as a
              result of this License or out of the use or inability to use the
              Work (including but not limited to damages for loss of goodwill,
              work stoppage, computer failure or malfunction, or any and all
              other commercial damages or losses), even if such Contributor
              has been advised of the possibility of such damages.
        
           9. Accepting Warranty or Additional Liability. While redistributing
              the Work or Derivative Works thereof, You may choose to offer,
              and charge a fee for, acceptance of support, warranty, indemnity,
              or other liability obligations and/or rights consistent with this
              License. However, in accepting such obligations, You may act only
              on Your own behalf and on Your sole responsibility, not on behalf
              of any other Contributor, and only if You agree to indemnify,
              defend, and hold each Contributor harmless for any liability
              incurred by, or claims asserted against, such Contributor by reason
              of your accepting any such warranty or additional liability.
        
           END OF TERMS AND CONDITIONS
        
           APPENDIX: How to apply the Apache License to your work.
        
              To apply the Apache License to your work, attach the following
              boilerplate notice, with the fields enclosed by brackets "[]"
              replaced with your own identifying information. (Don't include
              the brackets!)  The text should be enclosed in the appropriate
              comment syntax for the file format. We also recommend that a
              file or class name and description of purpose be included on the
              same "printed page" as the copyright notice for easier
              identification within third-party archives.
        
           Copyright 2024-2026 Heidi Andersén
        
           Licensed under the Apache License, Version 2.0 (the "License");
           you may not use this file except in compliance with the License.
           You may obtain a copy of the License at
        
               http://www.apache.org/licenses/LICENSE-2.0
        
           Unless required by applicable law or agreed to in writing, software
           distributed under the License is distributed on an "AS IS" BASIS,
           WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
           See the License for the specific language governing permissions and
           limitations under the License.
        
Project-URL: Homepage, https://github.com/heidihelena/epinet
Keywords: network-analysis,graph,centroid-clustering,epidemiology,quality-indicators
Requires-Python: >=3.10
Description-Content-Type: text/markdown
License-File: LICENSE
License-File: NOTICE
Requires-Dist: networkx>=2.5
Requires-Dist: numpy>=1.19.2
Requires-Dist: pandas>=1.1.5
Requires-Dist: scikit-learn>=0.24.1
Requires-Dist: matplotlib>=3.3.2
Requires-Dist: pyyaml>=5.4
Provides-Extra: lidc
Requires-Dist: pylidc>=0.2.3; extra == "lidc"
Provides-Extra: excel
Requires-Dist: xlrd>=2.0; extra == "excel"
Requires-Dist: openpyxl>=3.0; extra == "excel"
Provides-Extra: ui
Requires-Dist: streamlit>=1.30; extra == "ui"
Provides-Extra: xgboost
Requires-Dist: xgboost>=1.7; extra == "xgboost"
Provides-Extra: torch
Requires-Dist: torch>=2.0; extra == "torch"
Provides-Extra: dev
Requires-Dist: pytest>=7.0; extra == "dev"
Requires-Dist: ruff>=0.4; extra == "dev"
Requires-Dist: hypothesis>=6.0; extra == "dev"
Dynamic: license-file

# EpiNet — the Epistemic Network toolkit

[![Tests](https://github.com/heidihelena/epinet/actions/workflows/tests.yml/badge.svg)](https://github.com/heidihelena/epinet/actions/workflows/tests.yml)
[![License: Apache 2.0](https://img.shields.io/badge/License-Apache_2.0-blue.svg)](LICENSE)
![Python](https://img.shields.io/badge/python-3.10%E2%80%933.12-blue)
[![DOI](https://zenodo.org/badge/DOI/10.5281/zenodo.20681072.svg)](https://doi.org/10.5281/zenodo.20681072)

EpiNet is Vahtian’s research-methods demonstrator for graph-shaped biomedical data.
It produces calibrated, caveated, provenance-rich analysis reports. It is not clinical decision support.
CiteVahti checks claim-evidence support in manuscripts.
EpiNet checks whether graph-shaped analytical claims survive basic methodological stress tests.
Both share the same brand principles: human review, provenance, caveats, and no unsupported claims.

EpiNet is a **transparent node/edge network and feature-space analysis toolkit**
for graph-shaped datasets. The name reads *Epistemic Network*: the core question
is not just "what does the model predict" but how well-founded each call is — how
contestable, how calibrated, how well it transports. You load entities and
relationships from CSVs and it computes graph features, *honestly* evaluates an
outcome model, finds shortest paths, clusters nodes by feature-space centroid,
and scores how **contestable** each call is — with publication-quality figures
and a model card. Epidemiology is one use case; the core is domain-neutral
(driven through lung-cancer quality indicators, lung-nodule risk, and lymphoma
subtyping).

> **Scope.** This is a **research and education demonstrator, not clinical or
> public-health decision support.** Any model it produces must be validated on
> independent, outcome-linked data before it means anything clinically. See
> [Scope and caveats](#scope-and-caveats).

What distinguishes EpiNet from a thin scikit-learn wrapper is that **honest
evaluation is the default path**: calibration and uncertainty are reported with
the headline metric, while label-permutation nulls and community-aware splitting
are built into the same workflow when requested (and enabled by the Workbench's
default plan), so a good score can be checked against leakage or chance.
Developed as part of **Vahtian**; Apache-2.0 licensed.

**Jump to:** [About](#about-this-project) · [What it does](#what-it-does) ·
[Install](#install) · [Quick start](#quick-start) ·
[What it looks like](#what-it-looks-like) ·
[EpiNet Workbench](#epinet-workbench) · [Documentation](#documentation) ·
[Scope and caveats](#scope-and-caveats) · [Tests](#tests-and-linting) ·
[Citation](#citation) · [License](#license)

## About this project

EpiNet began a few years ago as a personal Python coding project and has grown
into a research-methods demonstrator with one organizing principle: **honest
evaluation is the default, not an afterthought.** Where most modelling code
optimizes a headline number, EpiNet is built to resist fooling itself — a
label-permutation null, calibration, bootstrap intervals, and community-aware
splitting live in the same run record rather than separate notebooks;
contestability makes each call inspectable; and the optional federated layer
shares findings across sites without pooling records. The name reads as
*Epistemic Network*: the question is not just what the model predicts, but how
well-founded each call is.

EpiNet's contribution is **methodological reproducibility, not a new predictor**:
the default model is a standard random forest, with logistic regression and
optional XGBoost alternatives, and the value is the conservative, auditable
evaluation workflow wrapped around the estimator so the same checks travel with
every analysis. It remains a **research and education demonstrator — not
clinical decision support.**

## What it looks like

<details>
<summary><b>Screenshot — the contestability lens on a lymphoma cohort</b></summary>

![Contestability panel on a lymphoma cohort](examples/sample-outputs/lymphoma/contestability.png)

*The contestability lens on the lymphoma grey-zone example. Left: how far each
call is from flipping, with the most-contested tail shaded. Right: value of
information — the features that most drive boundary flips (CD10, cyclin D1,
Ki67, …). The same lens runs on any cohort. More figures in
[examples/sample-outputs/](examples/sample-outputs/).*

</details>

## What it does

- **Graph features** — degree, weighted degree, clustering, component size,
  isolate flag, optional betweenness/closeness/PageRank.
- **Honest outcome model** — selectable estimator over graph features + node
  attributes: RandomForest by default, scaled regularized logistic regression as
  an interpretable comparator, optional XGBoost (`pip install
  "vahtian-epinet[xgboost]"`), or an optional tiny PyTorch MLP (`pip install
  "vahtian-epinet[torch]"`). Every option uses the same evaluation harness:
  discrimination (AUROC, AUPRC), classification (balanced accuracy, MCC, F1),
  **calibration** (Brier always; slope/intercept for binary outcomes), bootstrap
  CIs, permutation importance, a label-permutation null, community-aware
  splitting, small-cohort warnings, a reproducibility `provenance` block, and a
  TRIPOD+AI-flavoured `model_card.md`.
- **Shortest paths** — from sources to target nodes or target-outcome nodes, with
  per-target coverage.
- **Feature-space clustering** — k-means centroids + per-node distance to each
  outcome-class centroid (Euclidean or Ledoit–Wolf-shrunk Mahalanobis).
- **Contestability** (`--run-contest`) — the closed-form smallest feature-space
  move that flips a node's nearest-centroid class, plus a per-feature
  value-of-information ranking. See [docs/methods.md](docs/methods.md).
- **Input normalization** — maps common column aliases onto the schema before
  validation; never silently (every rename logged, raw + normalized hashed).
- **Epistemic claim gates** — graph semantics and timing are recorded so a
  similarity graph is not mistaken for observed social, clinical, or exposure
  relations. See [docs/epistemic-science.md](docs/epistemic-science.md).
- **Federated pipeline** (*optional*) — reconstruct the scaler, centroids, and
  contestability from per-site aggregates only, behind a fail-closed governance
  gate. See
  [docs/federated.md](docs/federated.md) and
  [docs/governance-and-consent.md](docs/governance-and-consent.md).
- **Baselines & external validation** — compare graph features against a
  node-embedding baseline and a no-information floor under the same harness, and
  validate a model on an independent cohort. See [docs/validation.md](docs/validation.md).
- **LLMvahti** (*experimental*) — the same contestability lens pointed at
  LLM-as-judge verdicts: a blinded-second-rater audit where the human rates
  first, with bootstrapped inter-rater agreement, judge-confidence calibration,
  and per-verdict flip-distance over rubric criteria. Run it with the
  `epinet-llmvahti` command (`--human`/`--judge` CSVs); see
  [docs/llmvahti.md](docs/llmvahti.md).

## Install

```bash
pip install vahtian-epinet   # from PyPI: the package + the `epinet` command
```

From a source checkout (for development or the example extras):

```bash
pip install -e .            # installs the package + the `epinet` command
pip install -e ".[dev]"     # also pytest + ruff + hypothesis (for development)
pip install -e ".[lidc]"    # pylidc, for the LIDC-IDRI / LUNA16 examples
pip install -e ".[excel]"   # xlrd + openpyxl, for the TCIA diagnosis spreadsheets
pip install -e ".[xgboost]" # optional XGBoost estimator backend
pip install -e ".[torch]"   # optional tiny PyTorch MLP estimator backend
```

`requirements.txt` lists the core runtime dependencies if you prefer not to
install the package.

## Quick start

```bash
epinet \
  --nodes synthetic_nodes.csv \
  --edges synthetic_edges.csv \
  --outcome-column Outcome \
  --target-outcome 1 \
  --model random_forest \
  --output-dir epinet_outputs
```

(`epinet ...` is the installed console command; `python -m vahtian.epinet.toolkit ...`
works identically without installing.) This runs graph-feature generation, an
honestly-evaluated outcome model, and shortest-path summaries side by side.

Key outputs in `epinet_outputs/`:

- `model_metrics.json` — discrimination, classification, calibration,
  `iteration_summary`, bootstrap CI, permutation test, data warnings, provenance
- `model_card.md` — TRIPOD+AI-flavoured human-readable model card
- `model_feature_importance.csv` — permutation importance (± `importance_std`)
- `node_features.csv`, `shortest_paths.csv`, `nearest_targets.csv`,
  `target_coverage.csv`, `provenance.json`, `run_summary.json`
- `plots/*.png` — network, calibration, learning curve, metric stability,
  confusion matrix, and more (see [docs/methods.md](docs/methods.md))

The data format is documented in [Data-format.md](Data-format.md).

## EpiNet Workbench

<details>
<summary><b>Local CSV-to-report interface — plan → run → optional UI</b></summary>

EpiNet Workbench is a local CSV-to-report interface. It does **not** replace the
command line and it is **not** AutoML: it never searches models or chases a
metric. It writes an `analysis.yaml` file and runs the *same* EpiNet engine as
the CLI, so every interactive run is reproducible without the interface.

```bash
# 1. Plan: profile the data, infer a schema, write a reviewable config
epinet-workbench plan --nodes synthetic_nodes.csv --edges synthetic_edges.csv \
  --outcome Outcome --output analysis.yaml

# 2. Run: execute the plan and write a portable result bundle
epinet-workbench run --config analysis.yaml

# 3. UI (optional): the same plan/run behind a local Streamlit workbench
pip install -e ".[ui]"
epinet-workbench ui
```

The config — not hidden UI state — is the source of truth. The five-screen UI
(Data → Schema → Plan → Run → Report) only ever builds an `analysis.yaml` and
calls the same runner.

**Supported modes**

1. **Single CSV** — feature-space analysis for an ordinary table (no network
   inference claimed unless you opt into a similarity graph).
2. **Nodes + edges** — the full graph pipeline on real graph-shaped data.
3. **Development + validation** — runs external validation by default; the mode
   for publishable work.

**Safety gates** block runs that would produce nonsense (no/single-class
outcome, ID used as a feature) and warn on the rest (suspected outcome leakage
selected as a feature, missing validation cohort, identifier-looking columns). A
too-small positive class downgrades model training to a descriptive report
rather than fabricating metrics.

**Scientific claims check.** Every run distils its diagnostics into
plain-language claim gates — written into the model card, `claims_check.json`,
and the HTML report:

- **Permutation null** — *signal above null* vs *signal not detected*.
- **Split sensitivity** — random vs community-aware split, to expose a headline
  that leaned on leakage between connected cases.
- **Baseline floor** — does the model beat the no-information baseline? The
  margin is measured *paired per split* (model and floor share the same splits),
  with a Nadeau–Bengio-corrected interval, so the verdict is three-way: *beats
  floor*, *at floor*, or *not resolvable at this n* when the interval straddles
  the line and the data cannot yet say.
- **External validation** — run or not, and how far performance transported.
- **Graph semantics** — whether the edge definition and edge timing license a
  graph-shaped predictive or epidemiological claim.
- A standing *"do not claim clinical utility unless…"* caveat, generated into
  every report and not removable by theming.

**Branded HTML report.** `epinet-workbench run` also writes a self-contained,
offline `index.html` into the bundle (summary, caveats, claims check, model card,
metrics, calibration, permutation null, baselines, contestability, plots,
provenance, CSV downloads) — the portable artifact to share or print to PDF. A
theme block in `analysis.yaml` sets the brand, title, logo, colours, and plot
palette (slide between the colourblind-safe **Wong** and the Vahtian **Sentinel**
palettes) — but the caveats, claims check, and provenance are always rendered:

```yaml
reporting:
  brand_name: "Vahtian / EpiNet"
  report_title: "EpiNet Analysis Report"
  logo_path: null
  primary_color: "#5E4F99"
  accent_color: "#8273C0"
  plot_palette: "wong"   # or "vahtian"
```

**Outputs** (the downloadable result bundle):
`index.html`, `analysis.yaml`, `model_metrics.json`, `model_card.md`,
`claims_check.json`, `split_comparison.json`, `provenance.json`,
`node_features.csv`, `node_contestability.csv`, `model_feature_importance.csv`,
`baseline_comparison.csv`, `external_validation.json` (publication mode),
publication-quality `plots/`, and `environment.txt`.

**Scope:** research and education only. Not clinical decision support.

</details>

## Documentation

<details>
<summary><b>Docs index — methods, examples, federated, governance, validation</b></summary>

- **[docs/methods.md](docs/methods.md)** — evaluation design (iterative
  evaluation, permutation null, community-aware splitting), the diagnostic
  figures, the contestability theory, and methodological boundaries.
- **[docs/examples.md](docs/examples.md)** — worked examples: shortest paths,
  the CiteMatch evidence graph, feature-space clustering, the pulmonary-nodule
  cohort, real LIDC-IDRI, and the Nordic lung-cancer quality-indicator network.
- **[docs/federated.md](docs/federated.md)** — the federated fit, federated
  contestability, the registry adapter, and the sealed-egress model.
- **[docs/governance-and-consent.md](docs/governance-and-consent.md)** — what the
  governance gate enforces vs what remains a policy/legal responsibility
  (explicitly non-legal).
- **[docs/validation.md](docs/validation.md)** — representation baselines (incl. a
  node-embedding comparison) and external validation: does the model transport?

Each worked example also has a builder script and a walkthrough under
`examples/*_usecase.md`; the federated and governance pipelines have runnable
demos under `examples/federated_*` and `examples/governance_*`.

</details>

## Scope and caveats

<details>
<summary><b>What EpiNet does <i>not</i> do, and what to add before any real-world use</b></summary>

The model is intentionally simple. It does **not** infer causality, outbreak
dynamics, clinical risk, or intervention effects. Network features can be useful
descriptors, but they can also encode sampling bias, measurement bias, and
structural confounding. **Use the outputs as exploratory evidence, not as
decisions.**

Before using EpiNet for health, education, welfare, employment, or public-sector
decisions, add: domain-specific data validation; directed/temporal assumptions;
uncertainty and sensitivity checks; external validation on independent
outcome-linked data; privacy and governance review; and human review of any
operational recommendation. The bundled cohorts are synthetic or small and
selection-biased — see the per-example limits in [docs/examples.md](docs/examples.md).
For clinical prediction, align reporting with
[TRIPOD+AI](https://doi.org/10.1136/bmj-2023-078378); for AI interventions, the
bar moves toward prospective evaluation (e.g. CONSORT-AI).

</details>

## Tests and linting

<details>
<summary><b>How to run the tests and the linter</b></summary>

```bash
python -m unittest discover -s tests   # or: pytest  (adds the hypothesis property tests)
ruff check .
```

GitHub Actions runs both on every push and pull request across Python
3.10–3.12 (`.github/workflows/tests.yml`).

</details>

## Citation

<details>
<summary><b>How to cite EpiNet</b></summary>

If you use EpiNet, please cite it via [`CITATION.cff`](CITATION.cff) (GitHub's
"Cite this repository" button generates APA/BibTeX from it).

</details>

## License

Apache License 2.0. See `LICENSE` and `NOTICE`. (Releases up to and including
`0.4.2` were published under the MIT License and remain available under those
terms; Apache-2.0 applies from `0.4.3` onward.)
