pyproject.toml

[build-system]
requires = ["setuptools>=68", "wheel"]
build-backend = "setuptools.build_meta"

[project]
name = "auto-cleaner"
version = "1.0.0"
description = "Autonomous, polars-native data preprocessing & automated EDA for ML-ready data."
readme = "README.md"
requires-python = ">=3.10"
license = { text = "MIT" }
authors = [{ name = "auto_cleaner" }]
keywords = ["polars", "duckdb", "data-cleaning", "eda", "preprocessing", "quant"]
classifiers = [
    "Programming Language :: Python :: 3 :: Only",
    "Intended Audience :: Science/Research",
    "Topic :: Scientific/Engineering",
]

# Core runtime deps. polars + numpy are mandatory; the rest unlock extra paths.
dependencies = [
    "polars>=1.0",
    "numpy>=1.24",
]

[project.optional-dependencies]
sql = ["duckdb>=0.10"]          # SQL ingestion via DuckDB
excel = ["fastexcel>=0.10"]     # Excel (.xlsx/.xlsm/.xls) ingestion via calamine
ml = ["scikit-learn>=1.3"]      # Isolation Forest + KNN imputation
parquet = ["pyarrow>=14"]       # Arrow interchange for SQL/Parquet
# kaleido 1.x needs a one-time headless-Chrome install: `plotly_get_chrome`
viz = ["plotly>=5.20", "kaleido>=1.0"]   # interactive charts + PNG export
stats = ["scipy>=1.10"]                  # normality tests + power transforms
inference = ["statsmodels>=0.14"]        # OLS/logit regression with p-values + CIs
astro = ["astropy>=5.0"]                 # FITS ingestion
climate = ["xarray>=2023.1", "netCDF4>=1.6"]  # netCDF gridded ingestion
advstats = ["pingouin>=0.5", "lifelines>=0.27", "pymannkendall>=1.4", "umap-learn>=0.5"]
nlp = ["vaderSentiment>=3.3"]            # classical sentiment
embed = ["sentence-transformers>=2.2"]  # opt-in neural embeddings (heavy)
pdf = ["fpdf2>=2.7"]                     # per-dataset PDF report
all = [
    "duckdb>=0.10", "fastexcel>=0.10", "scikit-learn>=1.3", "pyarrow>=14",
    "plotly>=5.20", "kaleido>=1.0",
    "scipy>=1.10", "statsmodels>=0.14", "astropy>=5.0", "xarray>=2023.1", "netCDF4>=1.6",
    "pingouin>=0.5", "lifelines>=0.27", "pymannkendall>=1.4", "umap-learn>=0.5",
    "vaderSentiment>=3.3", "fpdf2>=2.7",
]
dev = [
    "pytest>=7", "duckdb>=0.10", "fastexcel>=0.10", "xlsxwriter>=3.0",
    "scikit-learn>=1.3", "pyarrow>=14", "plotly>=5.20",
    "kaleido>=1.0", "scipy>=1.10", "statsmodels>=0.14", "astropy>=5.0", "xarray>=2023.1",
    "netCDF4>=1.6", "pingouin>=0.5", "lifelines>=0.27", "pymannkendall>=1.4",
    "umap-learn>=0.5", "vaderSentiment>=3.3", "fpdf2>=2.7",
]

[project.scripts]
auto-cleaner = "auto_cleaner.__main__:main"
auto-cleaner-predict = "auto_cleaner.predict:main"

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

[tool.pytest.ini_options]
testpaths = ["tests"]
addopts = "-q"