pyproject.toml

[build-system]
requires = ["setuptools>=61"]
build-backend = "setuptools.build_meta"

[project]
name = "aew"
version = "0.0.1"
description = "Reproducible pipeline for African Easterly Wave / convective-system analysis (Python port of the PhD NCL code)"
readme = "README.md"
license = "CC0-1.0"
requires-python = ">=3.10"
dependencies = [
  "numpy>=1.24",
  "scipy>=1.10",
  "xarray>=2023.1",
  "pandas>=2.0",
  "netCDF4>=1.6",
  "cftime>=1.6",
  # aew.v1port.detection contours the advection field with contourpy. It arrives as a
  # matplotlib transitive in most environments, which is exactly why it has to be declared:
  # a public clone that installs only the core would resolve it privately or not at all.
  "contourpy>=1.0",
]

[project.optional-dependencies]
plot = ["matplotlib>=3.7", "cartopy>=0.22"]
data = ["cdsapi>=0.6"]
stats = ["statsmodels>=0.14"]  # control_model.py (Poisson GLM, cluster-robust SE)
dev = ["pytest>=7", "pytest-cov>=4"]
# Everything scripts/run_canonical.py needs beyond the library core. The extras existed
# but nothing named the set the canonical sequence requires, so the documented
# "pip install -e ." could not run it. requirements-lock.txt pins the
# exact versions that produced the deposited record.
canonical = ["matplotlib>=3.8", "cartopy>=0.22", "statsmodels>=0.14"]

[tool.setuptools.packages.find]
where = ["src"]

[tool.pytest.ini_options]
testpaths = ["tests"]
filterwarnings = [
  "ignore:numpy.ndarray size changed:RuntimeWarning",
  "ignore:Setting the shape on a NumPy array:DeprecationWarning",
]