ggstyle 0.1.0__tar.gz

ggstyle-0.1.0/.gitignore

@@ -0,0 +1,16 @@
+.DS_Store
+.venv/
+__pycache__/
+*.py[cod]
+*.egg-info/
+.pytest_cache/
+.ruff_cache/
+.mypy_cache/
+.coverage
+htmlcov/
+build/
+dist/
+docs/build/
+docs/source/generated/
+*.tar.gz
+.release-inputs/

ggstyle-0.1.0/CHANGELOG.md

@@ -0,0 +1,23 @@
+# Changelog
+
+This project follows [Semantic Versioning](https://semver.org/).
+
+## 0.1.0 - 2026-08-19
+
+Initial public release.
+
+- Add a date-axis handle that adopts existing matplotlib axes.
+- Add independent tick cadence, label formatting, range, and grid controls.
+- Add observation-based gap collapsing and date-space annotations.
+- Add pandas, polars, pyarrow, NumPy, and plain-sequence date extraction.
+- Add opt-in minimal and grey themes without import-time global state changes.
+- Add NumPy-style API documentation and a warning-free Sphinx user guide modeled on the
+  documentation structure used by statsmodels.
+- Replace loosely typed annotation dictionaries with explicit internal state objects and
+  broaden automated clean-code checks.
+- Add structured ``AxisSummary`` metadata and captions generated from the same source.
+- Add explicit ``missing="raise"`` and ``missing="drop"`` date policies.
+- Add ``sync_dates`` for comparable date coordinates and limits across multiple panels.
+
+Collapsed mode currently remaps line artists only. Collection remapping remains planned
+for a later release.

ggstyle-0.1.0/CONTRIBUTING.md

@@ -0,0 +1,44 @@
+# Contributing
+
+Bug reports and focused pull requests are welcome. For behavior changes, open an issue
+first so the public API and date-axis semantics can be agreed before implementation.
+
+## Development
+
+Use Python 3.10 or newer:
+
+```bash
+python -m venv .venv
+.venv/bin/python -m pip install -e ".[dev]"
+.venv/bin/python -m pytest -q --cov=ggstyle --cov-report=term-missing
+.venv/bin/ruff check .
+.venv/bin/mypy src
+.venv/bin/python tools/validate_docstrings.py
+.venv/bin/python -m sphinx -W --keep-going -b html docs/source docs/build/html
+```
+
+Public functions, classes, methods, and attributes use the NumPy docstring standard, as
+in statsmodels. New functionality also belongs in the appropriate page under
+`docs/source` and in the changelog. Documentation must build without warnings.
+
+New behavior needs tests. Changes to collapsed coordinates should test both collapsed and
+expanded modes, including switching between them. Do not commit generated distributions,
+virtual environments, caches, or platform metadata.
+
+By contributing, you agree that your contributions are licensed under the MIT License.
+
+## Releasing
+
+Releases use PyPI trusted publishing; maintainers must not store a long-lived PyPI token
+in GitHub. Before the first release, register a pending publisher for project ``ggstyle``
+on PyPI with these values:
+
+- Owner: ``joshuamyers22``
+- Repository: ``ggstyle``
+- Workflow: ``publish.yml``
+- Environment: ``pypi``
+
+After the release commit passes CI, create and push a tag matching the package version,
+for example ``v0.1.0``. The publish workflow independently repeats the test, type,
+documentation, and package checks; publishes the distributions to PyPI; and creates the
+GitHub release only after publication succeeds.

ggstyle-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Joshua Myers
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ggstyle-0.1.0/PKG-INFO

@@ -0,0 +1,268 @@
+Metadata-Version: 2.5
+Name: ggstyle
+Version: 0.1.0
+Summary: A date axis for matplotlib that is easy to use and manipulate
+Project-URL: Homepage, https://github.com/joshuamyers22/ggstyle
+Project-URL: Repository, https://github.com/joshuamyers22/ggstyle
+Project-URL: Issues, https://github.com/joshuamyers22/ggstyle/issues
+Project-URL: Changelog, https://github.com/joshuamyers22/ggstyle/blob/main/CHANGELOG.md
+Author-email: Joshua Myers <joshua.myers22@gmail.com>
+License-Expression: MIT
+License-File: LICENSE
+Keywords: dates,ggplot2,matplotlib,visualization
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Developers
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3 :: Only
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Programming Language :: Python :: 3.13
+Classifier: Topic :: Scientific/Engineering :: Visualization
+Requires-Python: >=3.10
+Requires-Dist: matplotlib>=3.7
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Provides-Extra: dev
+Requires-Dist: build; extra == 'dev'
+Requires-Dist: mypy; extra == 'dev'
+Requires-Dist: pandas-stubs; extra == 'dev'
+Requires-Dist: polars>=1.0; extra == 'dev'
+Requires-Dist: pytest-cov; extra == 'dev'
+Requires-Dist: pytest>=7.0; extra == 'dev'
+Requires-Dist: ruff; extra == 'dev'
+Requires-Dist: twine; extra == 'dev'
+Provides-Extra: docs
+Requires-Dist: numpydoc>=1.6; extra == 'docs'
+Requires-Dist: pydata-sphinx-theme>=0.15; extra == 'docs'
+Requires-Dist: sphinx>=7; extra == 'docs'
+Provides-Extra: polars
+Requires-Dist: polars>=1.0; extra == 'polars'
+Description-Content-Type: text/markdown
+
+# ggstyle
+
+A date axis for matplotlib that is easy to use and easy to manipulate.
+
+**v0.1 is the date axis plus themes.** No palettes module and no `line()` yet — those
+remain future additions once the axis ergonomics have real usage behind them.
+
+This is the first public release. The date-axis behavior is tested, but the project is
+still young and follows semantic versioning. See the [known limits](#known-limits-in-v01)
+before using collapsed mode in production.
+
+## Why
+
+Most of the pain in Python time-series plotting is not the grammar, it's the axis: ticks in
+the wrong places, labels rotated to hide the fact that there are too many of them, weekend
+gaps shredding an intraday chart, and annotation code that quietly puts your vertical line
+three days off. `ggstyle` fixes the axis first.
+
+## Install
+
+```bash
+pip install ggstyle
+```
+
+For development from a clone:
+
+```bash
+pip install -e ".[dev]"
+```
+
+## Use
+
+It adopts any Axes, including plots it never made:
+
+```python
+import matplotlib.pyplot as plt
+import ggstyle as gs
+
+gs.use_theme()                            # "minimal" is the default
+
+fig, ax = plt.subplots()
+ax.plot(df["date"], df["close"])          # plain matplotlib, seaborn, or df.plot()
+
+gs.dates(ax).ticks("quarterly").fmt("month-year").zoom("2020", "2022")
+```
+
+Configuration and drawing methods return the handle, so calls chain.
+
+Axis semantics are also available as structured data rather than only rendered output:
+
+```python
+summary = gs.dates(ax).summary()
+caption = gs.dates(ax).caption(add=True)
+```
+
+### Ticks — where they go
+
+```python
+.ticks("monthly")                   # daily | weekly | monthly | quarterly | yearly
+.ticks("month-end")                 # anchored: month-start, quarter-end, year-start, ...
+.ticks(every="3M")                  # any offset alias; legacy M/Q/Y/H accepted
+.ticks(n=6)                         # about six ticks, snapped to a natural cadence
+.ticks(at=["2020-01-01", "2021-07-01"])
+.ticks(major="yearly", minor="monthly")
+```
+
+Anchoring is not cosmetic: month-start vs. month-end is the difference between labels that
+line up with your observations and labels that float between them.
+
+### Labels — what they say
+
+```python
+.fmt("concise")      # default: year shown once, not on every label
+.fmt("month-year")   # Jun 2020
+.fmt("quarter")      # Q2 2020
+.fmt("year") / .fmt("month") / .fmt("day") / .fmt("iso") / .fmt("time")
+.fmt("%b '%y")       # any strftime string
+.fmt(lambda d: f"week {d.isocalendar().week}")
+```
+
+Changing the format never moves a tick, and changing the cadence never changes the format.
+That orthogonality is a test, not an aspiration.
+
+### Range
+
+Partial strings expand to whole periods, pandas-style:
+
+```python
+.zoom("2020", "2022")      # three complete years
+.zoom("2020-03", None)     # open-ended
+.zoom(last="6M")           # trailing window from the last observation, not from today
+.zoom(ytd=True)
+.pad(left="1M", right="1M")
+```
+
+### Gaps
+
+```python
+.collapse()   # unobserved dates get no space
+.expand()     # true datetime axis, gaps restored
+```
+
+Collapsed mode is defined by **the dates present in your data**, not by a holiday calendar.
+Anything not observed is not allocated space. That is correct for any market or region and
+needs no extra dependency. With several series, the axis uses the union of observed dates.
+
+### Annotation in date space
+
+Every one of these is correct in both modes — that is the whole point of the handle:
+
+```python
+.loc("2020-03-23")                    # -> axis position; the escape-hatch primitive
+.vline("2020-03-23", label="trough")
+.span("2020-02-19", "2020-03-23", label="drawdown")
+.spans(events_df, start="begin", end="end", label="name")
+.grid("yearly")                       # gridline cadence, independent of ticks
+```
+
+In collapsed mode a date that falls inside a gap (a Sunday, a holiday) is placed by linear
+interpolation between its neighbours. `loc(date, snap=True)` rounds to the nearest
+observation instead; `loc(date, strict=True)` raises if the date was never observed.
+
+### Escape hatch
+
+`.loc()` is the primitive that keeps raw matplotlib correct:
+
+```python
+handle = gs.dates(ax).collapse()
+ax.axvline(handle.loc("2020-03-23"))   # lands in the right place
+ax.set_xlim(handle.loc("2020-01"), handle.loc("2021-01"))
+```
+
+## Themes
+
+Two ship. `minimal` is the default; `grey` is the ggplot2 `theme_grey` analogue.
+
+```python
+gs.use_theme()             # minimal, process-wide
+gs.use_theme("grey")       # "gray" also accepted
+
+with gs.theme("grey"):     # scoped; restores every rcParam on exit
+    ...
+
+plt.style.use(gs.stylesheet())   # the .mplstyle on its own, no ggstyle import needed
+```
+
+Both spell out the same type scale, colour cycle, and layout, so switching changes the
+panel surface and nothing else — the same separation ggplot2 makes. The colour cycle is
+Okabe–Ito-derived and capped at eight; past eight, direct labelling or faceting is the
+right answer, not a ninth colour.
+
+Importing `ggstyle` never mutates `rcParams`. Theming is always something you ask for.
+
+Almost all of it is plain rcParams in a `.mplstyle` file, including spine removal
+(`axes.spines.left: False`), which an earlier draft of the design wrongly assumed needed
+Python.
+
+## Data frames
+
+pandas and polars both work, as do pyarrow arrays, numpy `datetime64`, and plain lists:
+
+```python
+gs.dates(ax, data=frame["date"])     # pandas Series, polars Series, or Index
+```
+
+Polars is detected by module name rather than imported, so installing `ggstyle` never
+pulls it in and pandas-only users pay nothing for the support. Timezone-aware input from
+either library is converted to UTC instants for positioning; display timezones stay a
+separate concern handled by `.tz()`.
+
+Two things are deliberately *not* guessed: a whole DataFrame passed where a column was
+meant, and a string column that might be dates. Both raise.
+
+Missing values in explicit date data also raise unless exclusion is requested with
+``missing="drop"``. The number excluded remains available through ``.summary()`` and in
+generated captions.
+
+## Multiple panels
+
+Synchronize comparable axes with a common observation registry and limits:
+
+```python
+handles = gs.sync_dates(axes, mode="collapse", limits="union")
+```
+
+This prevents the same date from receiving different ordinal positions in independently
+collapsed panels.
+
+## Design rules
+
+- The date axis is a standalone object, not a side effect of plotting.
+- Importing the package is inert; theming is opt-in.
+- Placement, labels, gridline cadence, and range are four independent knobs.
+- Fail loudly: a non-date axis raises, and mixed tz-aware/naive input raises rather than
+  guessing UTC.
+- Never resample or interpolate the data silently.
+- Never rotate tick labels by default. Rotation is a symptom of bad tick selection.
+
+## Known limits in v0.1
+
+- Collapsed mode remaps `Line2D` artists only. Collections (`fill_between`, `scatter`) are
+  not yet remapped; annotate through the handle instead.
+- Native `ax.axvline(timestamp)` is still wrong in collapsed mode — go through `.loc()`.
+  A registered matplotlib scale would remove that caveat and is the v0.2 candidate.
+- `.tz()` assumes naive data is UTC when converting for display.
+- No palettes module yet: the colour cycle lives in the stylesheets.
+
+## Tests
+
+```bash
+python -m pytest -q
+ruff check .
+mypy src
+```
+
+See [CONTRIBUTING.md](CONTRIBUTING.md) for the complete development workflow and
+[SECURITY.md](SECURITY.md) for vulnerability reporting.
+
+The structured documentation follows the same user-guide, API-reference, pitfalls, and
+release-note separation used by statsmodels. Build it locally with:
+
+```bash
+pip install -e ".[docs]"
+python -m sphinx -W --keep-going -b html docs/source docs/build/html
+```

ggstyle-0.1.0/README.md

@@ -0,0 +1,225 @@
+# ggstyle
+
+A date axis for matplotlib that is easy to use and easy to manipulate.
+
+**v0.1 is the date axis plus themes.** No palettes module and no `line()` yet — those
+remain future additions once the axis ergonomics have real usage behind them.
+
+This is the first public release. The date-axis behavior is tested, but the project is
+still young and follows semantic versioning. See the [known limits](#known-limits-in-v01)
+before using collapsed mode in production.
+
+## Why
+
+Most of the pain in Python time-series plotting is not the grammar, it's the axis: ticks in
+the wrong places, labels rotated to hide the fact that there are too many of them, weekend
+gaps shredding an intraday chart, and annotation code that quietly puts your vertical line
+three days off. `ggstyle` fixes the axis first.
+
+## Install
+
+```bash
+pip install ggstyle
+```
+
+For development from a clone:
+
+```bash
+pip install -e ".[dev]"
+```
+
+## Use
+
+It adopts any Axes, including plots it never made:
+
+```python
+import matplotlib.pyplot as plt
+import ggstyle as gs
+
+gs.use_theme()                            # "minimal" is the default
+
+fig, ax = plt.subplots()
+ax.plot(df["date"], df["close"])          # plain matplotlib, seaborn, or df.plot()
+
+gs.dates(ax).ticks("quarterly").fmt("month-year").zoom("2020", "2022")
+```
+
+Configuration and drawing methods return the handle, so calls chain.
+
+Axis semantics are also available as structured data rather than only rendered output:
+
+```python
+summary = gs.dates(ax).summary()
+caption = gs.dates(ax).caption(add=True)
+```
+
+### Ticks — where they go
+
+```python
+.ticks("monthly")                   # daily | weekly | monthly | quarterly | yearly
+.ticks("month-end")                 # anchored: month-start, quarter-end, year-start, ...
+.ticks(every="3M")                  # any offset alias; legacy M/Q/Y/H accepted
+.ticks(n=6)                         # about six ticks, snapped to a natural cadence
+.ticks(at=["2020-01-01", "2021-07-01"])
+.ticks(major="yearly", minor="monthly")
+```
+
+Anchoring is not cosmetic: month-start vs. month-end is the difference between labels that
+line up with your observations and labels that float between them.
+
+### Labels — what they say
+
+```python
+.fmt("concise")      # default: year shown once, not on every label
+.fmt("month-year")   # Jun 2020
+.fmt("quarter")      # Q2 2020
+.fmt("year") / .fmt("month") / .fmt("day") / .fmt("iso") / .fmt("time")
+.fmt("%b '%y")       # any strftime string
+.fmt(lambda d: f"week {d.isocalendar().week}")
+```
+
+Changing the format never moves a tick, and changing the cadence never changes the format.
+That orthogonality is a test, not an aspiration.
+
+### Range
+
+Partial strings expand to whole periods, pandas-style:
+
+```python
+.zoom("2020", "2022")      # three complete years
+.zoom("2020-03", None)     # open-ended
+.zoom(last="6M")           # trailing window from the last observation, not from today
+.zoom(ytd=True)
+.pad(left="1M", right="1M")
+```
+
+### Gaps
+
+```python
+.collapse()   # unobserved dates get no space
+.expand()     # true datetime axis, gaps restored
+```
+
+Collapsed mode is defined by **the dates present in your data**, not by a holiday calendar.
+Anything not observed is not allocated space. That is correct for any market or region and
+needs no extra dependency. With several series, the axis uses the union of observed dates.
+
+### Annotation in date space
+
+Every one of these is correct in both modes — that is the whole point of the handle:
+
+```python
+.loc("2020-03-23")                    # -> axis position; the escape-hatch primitive
+.vline("2020-03-23", label="trough")
+.span("2020-02-19", "2020-03-23", label="drawdown")
+.spans(events_df, start="begin", end="end", label="name")
+.grid("yearly")                       # gridline cadence, independent of ticks
+```
+
+In collapsed mode a date that falls inside a gap (a Sunday, a holiday) is placed by linear
+interpolation between its neighbours. `loc(date, snap=True)` rounds to the nearest
+observation instead; `loc(date, strict=True)` raises if the date was never observed.
+
+### Escape hatch
+
+`.loc()` is the primitive that keeps raw matplotlib correct:
+
+```python
+handle = gs.dates(ax).collapse()
+ax.axvline(handle.loc("2020-03-23"))   # lands in the right place
+ax.set_xlim(handle.loc("2020-01"), handle.loc("2021-01"))
+```
+
+## Themes
+
+Two ship. `minimal` is the default; `grey` is the ggplot2 `theme_grey` analogue.
+
+```python
+gs.use_theme()             # minimal, process-wide
+gs.use_theme("grey")       # "gray" also accepted
+
+with gs.theme("grey"):     # scoped; restores every rcParam on exit
+    ...
+
+plt.style.use(gs.stylesheet())   # the .mplstyle on its own, no ggstyle import needed
+```
+
+Both spell out the same type scale, colour cycle, and layout, so switching changes the
+panel surface and nothing else — the same separation ggplot2 makes. The colour cycle is
+Okabe–Ito-derived and capped at eight; past eight, direct labelling or faceting is the
+right answer, not a ninth colour.
+
+Importing `ggstyle` never mutates `rcParams`. Theming is always something you ask for.
+
+Almost all of it is plain rcParams in a `.mplstyle` file, including spine removal
+(`axes.spines.left: False`), which an earlier draft of the design wrongly assumed needed
+Python.
+
+## Data frames
+
+pandas and polars both work, as do pyarrow arrays, numpy `datetime64`, and plain lists:
+
+```python
+gs.dates(ax, data=frame["date"])     # pandas Series, polars Series, or Index
+```
+
+Polars is detected by module name rather than imported, so installing `ggstyle` never
+pulls it in and pandas-only users pay nothing for the support. Timezone-aware input from
+either library is converted to UTC instants for positioning; display timezones stay a
+separate concern handled by `.tz()`.
+
+Two things are deliberately *not* guessed: a whole DataFrame passed where a column was
+meant, and a string column that might be dates. Both raise.
+
+Missing values in explicit date data also raise unless exclusion is requested with
+``missing="drop"``. The number excluded remains available through ``.summary()`` and in
+generated captions.
+
+## Multiple panels
+
+Synchronize comparable axes with a common observation registry and limits:
+
+```python
+handles = gs.sync_dates(axes, mode="collapse", limits="union")
+```
+
+This prevents the same date from receiving different ordinal positions in independently
+collapsed panels.
+
+## Design rules
+
+- The date axis is a standalone object, not a side effect of plotting.
+- Importing the package is inert; theming is opt-in.
+- Placement, labels, gridline cadence, and range are four independent knobs.
+- Fail loudly: a non-date axis raises, and mixed tz-aware/naive input raises rather than
+  guessing UTC.
+- Never resample or interpolate the data silently.
+- Never rotate tick labels by default. Rotation is a symptom of bad tick selection.
+
+## Known limits in v0.1
+
+- Collapsed mode remaps `Line2D` artists only. Collections (`fill_between`, `scatter`) are
+  not yet remapped; annotate through the handle instead.
+- Native `ax.axvline(timestamp)` is still wrong in collapsed mode — go through `.loc()`.
+  A registered matplotlib scale would remove that caveat and is the v0.2 candidate.
+- `.tz()` assumes naive data is UTC when converting for display.
+- No palettes module yet: the colour cycle lives in the stylesheets.
+
+## Tests
+
+```bash
+python -m pytest -q
+ruff check .
+mypy src
+```
+
+See [CONTRIBUTING.md](CONTRIBUTING.md) for the complete development workflow and
+[SECURITY.md](SECURITY.md) for vulnerability reporting.
+
+The structured documentation follows the same user-guide, API-reference, pitfalls, and
+release-note separation used by statsmodels. Build it locally with:
+
+```bash
+pip install -e ".[docs]"
+python -m sphinx -W --keep-going -b html docs/source docs/build/html
+```

ggstyle-0.1.0/SECURITY.md

@@ -0,0 +1,9 @@
+# Security policy
+
+`ggstyle` is a local plotting library: it does not make network requests, execute input,
+or deserialize untrusted objects. Treat custom matplotlib formatters and user-provided
+callables as trusted Python code.
+
+Please report a suspected vulnerability privately to `joshua.myers22@gmail.com` rather
+than opening a public issue. Include the affected version, a minimal reproducer, and the
+impact. Security fixes are made on the latest public release only.

ggstyle-0.1.0/docs/source/api.rst

@@ -0,0 +1,33 @@
+API reference
+=============
+
+The top-level API contains the date-axis handle, cadence specification, and opt-in theme
+helpers. Internal modules are not compatibility guarantees.
+
+Date axes
+---------
+
+.. autofunction:: ggstyle.dates
+
+.. autoclass:: ggstyle.DateAxis
+   :members:
+
+.. autoclass:: ggstyle.AxisSummary
+   :members:
+
+.. autoclass:: ggstyle.Cadence
+   :members:
+
+.. autofunction:: ggstyle.sync_dates
+
+Themes
+------
+
+.. autofunction:: ggstyle.use_theme
+
+.. autoclass:: ggstyle.theme
+   :members:
+
+.. autofunction:: ggstyle.stylesheet
+
+.. autofunction:: ggstyle.available_themes

ggstyle-0.1.0/docs/source/conf.py

@@ -0,0 +1,39 @@
+"""Sphinx configuration for the ggstyle documentation."""
+
+from __future__ import annotations
+
+from importlib.metadata import version as metadata_version
+
+project = "ggstyle"
+author = "Joshua Myers"
+copyright = "2026, Joshua Myers"
+release = metadata_version("ggstyle")
+
+extensions = [
+    "sphinx.ext.autodoc",
+    "sphinx.ext.doctest",
+    "sphinx.ext.viewcode",
+    "numpydoc",
+]
+
+autodoc_typehints = "none"
+numpydoc_class_members_toctree = False
+numpydoc_show_class_members = False
+numpydoc_xref_param_type = False
+
+doctest_global_setup = """
+from ggstyle import Cadence, available_themes, dates, use_theme
+"""
+doctest_global_cleanup = """
+import matplotlib.pyplot as plt
+plt.close("all")
+"""
+
+html_theme = "pydata_sphinx_theme"
+html_title = f"ggstyle {release}"
+html_theme_options = {
+    "github_url": "https://github.com/joshuamyers22/ggstyle",
+    "show_toc_level": 2,
+}
+
+exclude_patterns = ["_build", "generated"]