polars2svg 0.1.0__tar.gz

polars2svg-0.1.0/.github/dependabot.yml

@@ -0,0 +1,14 @@
+version: 2
+updates:
+  # Python dependencies (pyproject.toml: core deps, extras, and the dev
+  # dependency-group).
+  - package-ecosystem: "pip"
+    directory: "/"
+    schedule:
+      interval: "weekly"
+
+  # GitHub Actions used by ci.yml / release.yml.
+  - package-ecosystem: "github-actions"
+    directory: "/"
+    schedule:
+      interval: "weekly"

polars2svg-0.1.0/.github/workflows/ci.yml

@@ -0,0 +1,96 @@
+name: CI
+
+# Clean-room install verification on Linux (PyPI must-do #9).
+# Every prior test run was macOS/arm64; this exercises the built wheel on
+# linux/amd64 inside a stock python:3.13-slim image — a genuine clean room
+# (no preinstalled project deps, no system Python packages) — to catch
+# platform-dependent path handling, the mlx import guard, and font-metric
+# non-determinism in the golden-image tests.
+
+on:
+  push:
+    branches: [main]
+  pull_request:
+  workflow_dispatch:
+
+jobs:
+  # Type-check the public surface (PyPI nice-to-have #1). The package ships a
+  # py.typed marker, so downstream checkers read our hints; this guards that the
+  # public signatures stay valid. The [tool.mypy] config in pyproject.toml relaxes
+  # the error codes that fire only on the library's internal dynamic idiom.
+  type-check:
+    name: mypy (public surface)
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - name: Install uv
+        run: pip install --no-cache-dir uv
+      - name: Run mypy
+        run: uvx mypy polars2svg
+
+  # Security/quality automation (PyPI nice-to-have #5).
+  security-and-quality:
+    name: bandit + pip-audit + ruff
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: astral-sh/setup-uv@v5
+
+      - name: bandit (static security scan)
+        # Findings are either fixed or annotated in-line with a justified
+        # `# nosec <code> - <reason>` (see SECURITY.md's threat model for why
+        # e.g. random.randint()-based SVG id scoping and the background=
+        # shape-descriptor XML parsing are not vulnerabilities here). A clean
+        # run means every remaining flag has been reviewed and accepted.
+        run: uvx bandit -r polars2svg/
+
+      - name: pip-audit (known-vulnerable dependency scan)
+        # Audits the full resolved dependency graph (every extra + the dev
+        # group), not just what a bare `pip install polars2svg` pulls in —
+        # broader coverage than METADATA alone would give.
+        run: |
+          uv export --no-hashes --all-extras --group dev -o /tmp/requirements-audit.txt
+          uvx pip-audit -r /tmp/requirements-audit.txt
+
+      - name: ruff (E9, F)
+        # Minimal rule set on purpose — see the [tool.ruff] comment in
+        # pyproject.toml. Catches syntax errors and pyflakes issues (undefined
+        # names, unused imports/variables, duplicate dict keys) without
+        # fighting the codebase's distinctive style.
+        run: uvx ruff check polars2svg/
+
+  linux-cleanroom:
+    name: Linux clean-room wheel install + tests
+    runs-on: ubuntu-latest
+    container: python:3.13-slim
+    steps:
+      # git lets actions/checkout use a fast clone instead of the REST tarball.
+      - name: Install git
+        run: apt-get update && apt-get install -y --no-install-recommends git
+
+      - uses: actions/checkout@v4
+
+      - name: Build the wheel
+        run: |
+          pip install --no-cache-dir uv
+          uv build
+
+      - name: Install the built wheel into a clean environment
+        run: pip install --no-cache-dir dist/*.whl
+
+      - name: Import smoke test
+        run: python -c "import polars2svg; print('import ok')"
+
+      - name: Install test-only dependencies
+        # Mirrors the [dependency-groups] dev set that the wheel METADATA does
+        # not carry: pytest + the rasterization/kaggle test extras. rlPyCairo is
+        # reportlab 5.x's renderPM backend on Linux (macOS has its own); without
+        # it svglib rasterization errors out. The PNG-RMS golden tests skip off
+        # macOS regardless, but the backend is needed for any that force-run.
+        run: pip install --no-cache-dir pytest kagglehub svglib reportlab rlPyCairo
+
+      - name: Run the test suite
+        # test_performance.py needs a machine-local baseline (gitignored);
+        # kaggle-backed t-field tests need network/credentials. Both excluded,
+        # matching must-do #9.
+        run: python -m pytest tests/ -q --ignore=tests/test_performance.py -k 'not tfield'

polars2svg-0.1.0/.github/workflows/docs.yml

@@ -0,0 +1,50 @@
+# Build the mkdocs site and deploy it to GitHub Pages.
+#
+# The site builds from static, committed content (the example SVGs are
+# generated locally by docs/gen_examples.py), so this job does not need
+# polars2svg installed. Once the package source is mirrored into this repo /
+# published to PyPI, add an install step here and an mkdocstrings API page.
+#
+# Requires: repo Settings → Pages → Source = "GitHub Actions".
+name: docs
+
+on:
+  push:
+    branches: [main]
+  workflow_dispatch:
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+concurrency:
+  group: pages
+  cancel-in-progress: true
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
+        with:
+          python-version: '3.13'
+      - name: Install mkdocs-material
+        run: pip install --no-cache-dir mkdocs-material
+      - name: Build site
+        run: mkdocs build --strict
+      - uses: actions/configure-pages@v5
+      - uses: actions/upload-pages-artifact@v3
+        with:
+          path: site
+
+  deploy:
+    needs: build
+    runs-on: ubuntu-latest
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
+    steps:
+      - id: deployment
+        uses: actions/deploy-pages@v4

polars2svg-0.1.0/.github/workflows/release.yml

@@ -0,0 +1,63 @@
+name: Release
+
+# Publishes polars2svg to PyPI via Trusted Publishing (OIDC) on release tags.
+#
+# This is NOT continuous integration: it runs only when a tag matching v*
+# is pushed (e.g. `git tag v0.1.0 && git push origin v0.1.0`), so it does
+# not run the test suite on every commit (see PyPI must-do #11 — CI-on-every-
+# commit is intentionally disabled; local/standalone-server runs cover testing).
+#
+# One-time setup the maintainer must do (an agent cannot): register this repo
+# as a Trusted Publisher on PyPI. Before the project exists, add it as a
+# "pending publisher" in your PyPI account:
+#   https://pypi.org/manage/account/publishing/
+# with workflow name `release.yml` and environment name `pypi`. This replaces
+# the twine + API-token flow that used to live in pypi_instructions.txt.
+
+on:
+  push:
+    tags: ['v*']
+  workflow_dispatch:
+
+jobs:
+  build:
+    name: Build sdist + wheel
+    runs-on: ubuntu-latest
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Install uv
+        uses: astral-sh/setup-uv@v5
+
+      - name: Build distributions
+        run: uv build
+
+      - name: Check distribution metadata
+        run: uvx twine check dist/*
+
+      - name: Upload distributions
+        uses: actions/upload-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+  publish:
+    name: Publish to PyPI
+    needs: build
+    runs-on: ubuntu-latest
+    # Trusted Publishing: no API token/secret. PyPI verifies the OIDC identity
+    # of this workflow running in the `pypi` environment.
+    environment:
+      name: pypi
+      url: https://pypi.org/project/polars2svg/
+    permissions:
+      id-token: write
+    steps:
+      - name: Download distributions
+        uses: actions/download-artifact@v4
+        with:
+          name: dist
+          path: dist/
+
+      - name: Publish to PyPI
+        uses: pypa/gh-action-pypi-publish@release/v1

polars2svg-0.1.0/.gitignore

@@ -0,0 +1,222 @@
+# Byte-compiled / optimized / DLL files
+__pycache__/
+*.py[codz]
+*$py.class
+
+# C extensions
+*.so
+
+# Distribution / packaging
+.Python
+build/
+develop-eggs/
+dist/
+downloads/
+eggs/
+.eggs/
+lib/
+lib64/
+parts/
+sdist/
+var/
+wheels/
+share/python-wheels/
+*.egg-info/
+.installed.cfg
+*.egg
+MANIFEST
+
+# PyInstaller
+#  Usually these files are written by a python script from a template
+#  before PyInstaller builds the exe, so as to inject date/other infos into it.
+*.manifest
+*.spec
+
+# Installer logs
+pip-log.txt
+pip-delete-this-directory.txt
+
+# Unit test / coverage reports
+htmlcov/
+.tox/
+.nox/
+.coverage
+.coverage.*
+.cache
+nosetests.xml
+coverage.xml
+*.cover
+*.py.cover
+.hypothesis/
+.pytest_cache/
+cover/
+coverage.json
+
+# Translations
+*.mo
+*.pot
+
+# Django stuff:
+*.log
+local_settings.py
+db.sqlite3
+db.sqlite3-journal
+
+# Flask stuff:
+instance/
+.webassets-cache
+
+# Scrapy stuff:
+.scrapy
+
+# Sphinx documentation
+docs/_build/
+
+# PyBuilder
+.pybuilder/
+target/
+
+# Jupyter Notebook
+.ipynb_checkpoints
+
+# IPython
+profile_default/
+ipython_config.py
+
+# pyenv
+#   For a library or package, you might want to ignore these files since the code is
+#   intended to run in multiple environments; otherwise, check them in:
+# .python-version
+
+# pipenv
+#   According to pypa/pipenv#598, it is recommended to include Pipfile.lock in version control.
+#   However, in case of collaboration, if having platform-specific dependencies or dependencies
+#   having no cross-platform support, pipenv may install dependencies that don't work, or not
+#   install all needed dependencies.
+#Pipfile.lock
+
+# UV
+#   Similar to Pipfile.lock, it is generally recommended to include uv.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#uv.lock
+#   uv.toml holds machine-local resolver settings (exclude-newer); keep it out of VCS.
+uv.toml
+
+# poetry
+#   Similar to Pipfile.lock, it is generally recommended to include poetry.lock in version control.
+#   This is especially recommended for binary packages to ensure reproducibility, and is more
+#   commonly ignored for libraries.
+#   https://python-poetry.org/docs/basic-usage/#commit-your-poetrylock-file-to-version-control
+#poetry.lock
+#poetry.toml
+
+# pdm
+#   Similar to Pipfile.lock, it is generally recommended to include pdm.lock in version control.
+#   pdm recommends including project-wide configuration in pdm.toml, but excluding .pdm-python.
+#   https://pdm-project.org/en/latest/usage/project/#working-with-version-control
+#pdm.lock
+#pdm.toml
+.pdm-python
+.pdm-build/
+
+# pixi
+#   Similar to Pipfile.lock, it is generally recommended to include pixi.lock in version control.
+#pixi.lock
+#   Pixi creates a virtual environment in the .pixi directory, just like venv module creates one
+#   in the .venv directory. It is recommended not to include this directory in version control.
+.pixi
+
+# PEP 582; used by e.g. github.com/David-OConnor/pyflow and github.com/pdm-project/pdm
+__pypackages__/
+
+# Celery stuff
+celerybeat-schedule
+celerybeat.pid
+
+# SageMath parsed files
+*.sage.py
+
+# Environments
+.env
+.envrc
+.venv
+env/
+venv/
+ENV/
+env.bak/
+venv.bak/
+
+# Spyder project settings
+.spyderproject
+.spyproject
+
+# Rope project settings
+.ropeproject
+
+# mkdocs documentation
+/site
+
+# mypy
+.mypy_cache/
+.dmypy.json
+dmypy.json
+
+# Pyre type checker
+.pyre/
+
+# pytype static type analyzer
+.pytype/
+
+# Cython debug symbols
+cython_debug/
+
+# PyCharm
+#  JetBrains specific template is maintained in a separate JetBrains.gitignore that can
+#  be found at https://github.com/github/gitignore/blob/main/Global/JetBrains.gitignore
+#  and can be added to the global gitignore or merged into this file.  For a more nuclear
+#  option (not recommended) you can uncomment the following to ignore the entire idea folder.
+#.idea/
+
+# Abstra
+# Abstra is an AI-powered process automation framework.
+# Ignore directories containing user credentials, local state, and settings.
+# Learn more at https://abstra.io/docs
+.abstra/
+
+# Visual Studio Code
+#  Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore
+#  that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
+#  and can be added to the global gitignore or merged into this file. However, if you prefer,
+#  you could uncomment the following to ignore the entire vscode folder
+# .vscode/
+
+# Ruff stuff:
+.ruff_cache/
+
+# PyPI configuration file
+.pypirc
+
+# Cursor
+#  Cursor is an AI-powered code editor. `.cursorignore` specifies files/directories to
+#  exclude from AI features like autocomplete and code analysis. Recommended for sensitive data
+#  refer to https://docs.cursor.com/context/ignore-files
+.cursorignore
+.cursorindexingignore
+
+# Marimo
+marimo/_static/
+marimo/_lsp/
+__marimo__/
+coverage_reports/20260212_2230_coverage.json
+
+# Claude
+.claude
+# CLAUDE.md
+# *.md
+
+# Performance baseline (machine-specific, auto-generated by test_performance.py)
+tests/perf_baseline.json
+
+graphify-out/
+.DS_Store

polars2svg-0.1.0/CHANGELOG.md

@@ -0,0 +1,130 @@
+# Changelog
+
+All notable changes to **polars2svg** are documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
+and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+### Added
+
+- **`TFDPLayout` now runs on NVIDIA GPUs via MLX's CUDA backend.** The layout was
+  previously documented and packaged as Apple-silicon-only. MLX ships an official
+  CUDA backend, and the eight t-FDP compute kernels are backend-agnostic
+  `mlx.core` — so the same code runs on Metal or CUDA with **no change to the
+  math**. New `[mlx-cuda]` / `[mlx-cuda13]` extras install it (Linux only; NVIDIA
+  SM ≥ 7.5, driver ≥ 550.54.14, glibc ≥ 2.35).
+- **`polars2svg.gpu_backend()`** — reports which backend `TFDPLayout` resolved to:
+  `'metal'`, `'cuda'`, or `'cpu'`.
+- **`tests/test_tfdp_backend.py`** — exercises each MLX op t-FDP depends on
+  (broadcast all-pairs diff, gather, `.at[].add()` scatter-add with duplicate
+  indices, keyed RNG) directly on the resolved device, so a backend gap fails by
+  name. Plus GPU-vs-CPU cross-check, convergence, and finiteness tests in
+  `test_tfdp_layout.py`.
+
+- **Legends and colorbars: opt-in `legend=` on every rendered component**
+  (`xyp`, `histop`, `timep`, `piep`, `linkp`, `chordp`, `spreadlinesp`;
+  `smallp` panels inherit it from their template component). Layered value:
+  `legend=True` (≡ `'right'`), a position string
+  (`'right' | 'left' | 'top' | 'bottom'`), or a dict
+  (`{'pos', 'title', 'fmt', 'max_items', 'order'}`). The legend **kind is
+  auto-selected** from the resolved color mode — a categorical swatch list
+  for `CSETp`/bare-categorical color, a colorbar for the spectrum modes —
+  and the strip is reserved **from** `wxh` (the plot region shrinks; the
+  physical output size is unchanged). A truthy `legend` with nothing to
+  legend (flat/fixed color) silently renders nothing, so
+  `set_defaults(legend=True)` is safe as a global default. Legends are
+  recorded into the shared `DisplayList`, so SVG **and** WebGPU outputs both
+  carry them, and the captured scale/category metadata is exposed as
+  `component.legend_info` (a `polars2svg.LegendInfo`). v1 scope is the color
+  encoding only (no size legends yet). The default `legend=False` renders
+  byte-identically to previous output.
+
+### Fixed
+
+- **`pip install polars2svg[mlx]` was broken.** `tfdp_layout.py` imports
+  `scipy`/`networkx`/`scikit-learn`, but the `mlx` extra did not chain `[layouts]`
+  (as `interactive` does), so the extra resolved to an install whose only module
+  could not import. `mlx` now chains `polars2svg[layouts]`.
+
+### Changed
+
+- `TFDPLayout` no longer hardcodes `device=mx.gpu`. It probes the GPU once and
+  falls back to the CPU device with a one-time warning if none is usable — the
+  plain Linux MLX wheel has no GPU backend, which would otherwise have failed at
+  the first kernel.
+
+- **Slimmed the core install; added `[layouts]` and `[interactive]` extras.**
+  `pip install polars2svg` now pulls only `polars`, `numpy`, `pyarrow`,
+  `pillow`, and `platformdirs` — enough for `xyp`, `histop`, `timep`, `piep`,
+  `smallp`, and `spreadlinesp`. `networkx`, `scikit-learn`, `scipy`, `shapely`,
+  and `squarify` move to the new `polars2svg[layouts]` extra (needed by
+  `linkp`'s/`chordp`'s pluggable layouts, the graph-layout mixin, and
+  shapely-typed `background=` shapes on `xyp`/`linkp`); `panel`,
+  `jupyter_bokeh`, and `param` move to `polars2svg[interactive]` (which
+  includes `layouts`, since the interactive graph views need it too).
+  `polars2svg[all]` restores the previous "batteries included" install.
+  Calling a component that needs a missing extra (`chordp()`, `panelize()`,
+  `xypi()`, a graph-layout mixin method, `background=` shapes) now raises a
+  clear `ImportError` naming the extra to install, rather than either
+  succeeding silently or failing with a bare `ModuleNotFoundError` — this
+  breaks nothing at runtime as long as the extra is installed, but is a
+  **breaking change to the default install** if you were relying on
+  `linkp`/`chordp`/interactive variants working out of the box with a bare
+  `pip install polars2svg`.
+
+## [0.1.0] — 2026-07-10
+
+Initial public release: Polars-native DataFrame → SVG visualizations for Jupyter.
+Pass a Polars `DataFrame` straight to a component and get a crisp, self-contained
+SVG back. Every component also has a linked, interactive variant (built on Panel)
+for brushing and cross-filtering in a notebook, plus optional WebGPU rendering for
+large frames.
+
+### Added
+
+- **Component roster** — eight rendered components, each a method on a
+  `Polars2SVG` instance:
+  - `xyp` — scatter / distribution plot (numeric or categorical x/y).
+  - `histop` — horizontal histogram bars, one per category/bin.
+  - `timep` — temporal bar chart with linear or periodic (day-of-week, month, …)
+    time modes.
+  - `linkp` — node-link graph / network with pluggable layouts.
+  - `chordp` — chord diagram of weighted flows around a circle.
+  - `piep` — pie chart.
+  - `spreadlinesp` — egocentric radial "spread" rings for influence/propagation.
+  - `smallp` — small multiples: a grid of one template component faceted by a field.
+- **Interactive, cross-linked variants** — `xypi`, `histopi`, `timepi`, `linkpi`,
+  `smallpi` (and peers) sharing the static signatures, composed into a dashboard
+  with `panelize(layout)` for brushing and cross-filtering in a notebook.
+- **WebGPU rendering** — pass `use_webgpu=True` to render through WebGPU for large
+  frames; a `webgpu()` / `gpuDisplayList()` path exists across all components.
+- **Orthogonal `count=` / `color=` encoding** — a shared aggregation rule for
+  `count=` (row count, sum, distinct-count, struct distinct-count) and a
+  dtype-keyed `color=` inference (numeric → magnitude spectrum, otherwise
+  categorical), with enums (`SCALARp`, `SETp`, `CSETp`, `CMAGNITUDE_SUMp`,
+  `CROW_MAGNITUDEp`, `CROW_STRETCHEDp`, …) to pin intent explicitly.
+- **T-fields** — `p2s.tField(column, enum)` for time-transformation fields
+  (returns a frozen `TField` `str` subclass); the legacy `'column|suffix'` string
+  form still works with a one-time deprecation warning.
+- **Pluggable layout classes** exported from the package: `PolarsForceDirectedLayout`,
+  `ConveyProximityLayout`, `LandmarkMDSLayout`, `PivotMDSLayout`, `TFDPLayout`
+  (all satisfy the `LayoutAlgorithm` protocol via `.results()`).
+- **Export API** — `save(path)` / `savePNG(path)` on every rendered component;
+  SVG save has no extra dependency, PNG rasterization is behind the `[export]`
+  extra (`svglib`, `reportlab`, `rlPyCairo`).
+- **Typed package** — ships a `py.typed` marker with type hints on the public
+  surface (constructor, component factories, `tField`, `panelize`, layout classes).
+- **Exception hierarchy** — `Polars2SVGError` base with `InvalidSpecError` and
+  `DataError` subclasses, all exported from the package.
+- **`__version__`** attribute, read from installed package metadata.
+- **Optional extras** — `[mlx]` (Apple-silicon MLX-accelerated force-directed
+  layout) and `[export]` (PNG rasterization).
+- **Diagnostic INFO logging** through the `polars2svg_logger` for dtype-keyed
+  `count=`/`color=` inference choices (off by default).
+- **SECURITY.md** documenting the SVG-injection threat model (row-data label text
+  is HTML-escaped; component configuration is trusted).
+
+[Unreleased]: https://github.com/datrcode/polars2svg/compare/v0.1.0...HEAD
+[0.1.0]: https://github.com/datrcode/polars2svg/releases/tag/v0.1.0