pyproject.toml

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

[project]
name = "tabpfn-extensions"
version = "0.6.3"
dependencies = [
    "torch>=2.1",
    "pandas>=2.2.2",
    "scikit-learn>=1.6.0",
    "scipy>=1.11.1",
    "tabpfn>=8.4.0",
    "tabpfn-common-utils[telemetry-interactive]>=0.2.13",
]

requires-python = ">=3.10"
authors = [
    { name = "Noah Hollmann", email = "noah.hollmann@charite.de" },
    { name = "Leo Grinsztajn" },
    { name = "Samuel Müller", email = "muellesa@cs.uni-freiburg.de" },
    { name = "Lennart Purucker" },
    { name = "Arjun Krishnakumar" },
    { name = "Max Körfer" },
    { name = "Shi Bin Hoo" },
    { name = "Robin Tibor Schirrmeister" },
    { name = "Frank Hutter", email = "fh@cs.uni-freiburg.de" },
    { name = "Eddie Bergman" },
    { name = "Klemens Flöge" },
    { name = "Oscar Key" },
    { name = "Dominik Safaric" }
]
readme = "README.md"
description = "TabPFN: Foundation model for tabular data"
classifiers = [
    'Intended Audience :: Science/Research',
    'Intended Audience :: Developers',
    'License :: OSI Approved :: Apache Software License',
    'Programming Language :: Python',
    'Topic :: Software Development',
    'Topic :: Scientific/Engineering',
    'Operating System :: POSIX',
    'Operating System :: Unix',
    'Operating System :: MacOS',
    'Programming Language :: Python :: 3',
    'Programming Language :: Python :: 3.10',
    'Programming Language :: Python :: 3.11',
    'Programming Language :: Python :: 3.12',
    'Programming Language :: Python :: 3.13',
]
license = { file = "LICENSE" }

[project.urls]
documentation = "https://priorlabs.ai/docs"
source = "https://github.com/PriorLabs/tabpfn-extensions"

[project.optional-dependencies]
interpretability = [
    # get_tabpfn_inf_explainer needs shapiq>=1.4.0 (MarginalImputer +
    # get_predict_function_and_model_type(..., class_index=...); 1.3.0 is broken
    # via an unpackaged `overrides` dep). OddSHAP/ProxySHAP routing (see
    # interpretability/shapiq.py) additionally needs shapiq>=1.6.0, which
    # requires Python>=3.12; on older Python the routing falls back to shapiq's
    # "auto" selection with a warning.
    "shapiq>=1.6.0; python_version >= '3.12'",
    "shapiq>=1.4.0; python_version <  '3.12'",
    # Surrogate backends: ProxySHAP uses XGBoost, OddSHAP LightGBM; without them
    # both fall back to a weaker DecisionTree. Only needed where those estimators
    # exist (shapiq>=1.6.0, i.e. Python>=3.12).
    "xgboost>=2.1.0; python_version >= '3.12'",
    # Capped below 4.7.0 to match the resolution-wide constraint in [tool.uv];
    # see the note there. Worth revisiting once that cap can be dropped.
    "lightgbm>=4.3.0,<4.8.0; python_version >= '3.12'",
    "seaborn>=0.12.2",
    # decoder_readout.plot_decoder_readout draws with matplotlib.
    "matplotlib>=3.8.4",
]
# Install scikit-survival manually if you need SurvivalTabPFN
# `pip install scikit-survival`. Excluded due to GPL.
survival = []
many_class = []
unsupervised = []
bayesian_optimization = [
    # Only needed for the benchmark objective in the example.
    "botorch>=0.12.0",
]
# transformers for DINOv3, torchvision for its preprocessing, pillow for image decoding.
image = [
    "transformers>=4.57.1",
    "torchvision>=0.20",
    "pillow>=10.4",
]

# Meta-package that installs all extensions
all = [
    # scikit-survival not included; install manually if you need SurvivalTabPFN.
    "tabpfn-extensions[interpretability,survival,bayesian_optimization,image]",
]

[dependency-groups]
# The minimum subset of the dev dependencies required to run the tests on the CI.
# The idea is to be as close to the deployment environment as possible.
ci = [
    "pytest>=8.0.0",
    "pytest-mock>=3.15.1",
    "licensecheck>=2025.1.0",
]
dev = [
    {include-group = "ci"},
    "pre-commit>=3.0.0",
    "ruff==0.8.6", # This must be the same version as in .pre-commit-config.yaml
    "mypy>=1.0.0",
    "pytest-xdist>=3.6.0",
    "build>=1.3.0",
    "twine>=6.2.0",
]
# Extra deps needed only to *run* the example scripts (see examples/), used by the
# example-file CI job. Deliberately not a published extra: the library itself does
# not need these, only the example demos do, and the examples ship in the repo (not
# the wheel), so anyone running them works from a clone where this group is available.
examples = [
    # Only used by interpretability/shap_example.py for plotting.
    # Floor is 0.46.0 (0.41.0 pins an unbuildable numba).
    "shap>=0.46.0",
    # Only used by interpretability/decoder_readout_example.py for the 2D
    # embedding projection; the example falls back to PCA when it is absent.
    "umap-learn>=0.5.0",
]

[tool.uv]
# lightgbm 4.7.0 switched to PEP 639 license metadata and PyPI reports no license
# at all for that release, so the CI license check sees an empty license and fails.
# tabpfn itself depends on lightgbm>=3.0, so the cap has to be a resolution-wide
# constraint; bounding it in the interpretability extra alone does not reach the
# transitive copy. Worth revisiting once PyPI reports its MIT license again.
constraint-dependencies = ["lightgbm<4.7.0"]
exclude-newer = "7 days"
exclude-newer-package = {torch = false, tabpfn = false}

[tool.licensecheck]
only_licenses = ["APACHE", "MIT", "BSD", "ISC", "PYTHON", "UNLICENSE", "UNKNOWN"]
ignore_packages = [
    "certifi",
    "fsspec",
    "tabpfn*",
]

[tool.pytest.ini_options]
testpaths = ["tests"]  # Where the tests are located
minversion = "8.0"
empty_parameter_set_mark = "xfail"  # Prevents user error of an empty `parametrize` of a test
log_cli = false
log_level = "DEBUG"
xfail_strict = true
addopts = "--durations=10 -vv"

# https://github.com/astral-sh/ruff
[tool.ruff]
target-version = "py310"
line-length = 88
output-format = "full"
src = ["src", "tests", "examples"]
# misc/sklearn_compat.py is vendored verbatim from sklearn-compat upstream;
# excluding it from both lint and format keeps it byte-identical to upstream
# so future re-vendors are clean diffs.
extend-exclude = ["src/tabpfn_extensions/misc/sklearn_compat.py"]

[tool.ruff.lint]
# Allow unused variables when underscore-prefixed.
dummy-variable-rgx = "^(_+|(_+[a-zA-Z0-9_]*[a-zA-Z0-9]+?))$"

select = [
    "A",
    # "ANN", # Handled by mypy
    "ARG",
    "B",
    "BLE",
    # COM812 removed to avoid conflicts with formatter
    "COM",
    "C4",
    "D",
    # "DTZ",  # One day I should know how to utilize timezones and dates...
    "E",
    # "EXE", Meh
    "ERA",
    "F",
    "FBT",
    "I",
    # "ISC",  # Favours implicit string concatenation
    "INP",
    # "INT", # I don't understand this one
    "N",
    "NPY",
    "PD",
    "PLC",
    "PLE",
    "PLR",
    "PLW",
    "PIE",
    "PT",
    "PTH",
    # "PYI", # Specific to .pyi files for type stubs
    "Q",
    "PGH004",
    "RET",
    "RUF",
    "C90",
    "S",
    # "SLF",    # Private member accessed (sure, it's python)
    "SIM",
    # "TRY", # Good in principle, would take a lot of work to statisfy
    "T10",
    "T20",
    "TID",
    # TC003 replaced TCH003
    "TC",
    "UP",
    "N",
    "W",
    "YTT",
]

ignore = [
    "D104", # Missing docstring in public package
    "D105", # Missing docstring in magic mthod
    "D203", # 1 blank line required before class docstring
    "D205", # 1 blank line between summary and description
    "D401", # First line of docstring should be in imperative mood
    "N806", # Variable X in function should be lowercase
    "E731", # Do not assign a lambda expression, use a def
    "A002", # Shadowing a builtin
    "A003", # Shadowing a builtin
    "S101", # Use of assert detected.
    "W292", # No newline at end of file
    "PLC1901", # "" can be simplified to be falsey
    "TC003", # Move stdlib import into TYPE_CHECKING
    "COM812", # Missing trailing comma in imports
    "PLR2004", # Magic numbers, gets in the way a lot
    "PLR0915", # Too many statements
    "N803", # Argument name `X` should be lowercase
    "N802", # Function name should be lowercase
    "PLR0913", # Too many arguments in function definition
    "SIM102", # Use a single if statement instead of nested if
    "B007", # Loop control variable not used within loop body
    "N815", # Variable should not be mixedCase
    "S311", # Standard pseudo-random generators not suitable for crypto
    "RUF010", # Use explicit conversion flag
    "RET504", # Unnecessary assignment before return
    "RUF009", # Do not perform function call in dataclass defaults
    "SIM105", # Use contextlib.suppress instead of try-except-pass
    "PLW2901", # For loop variable overwritten by assignment target
    "SIM114", # Combine if branches using logical or
    "PLW0603", # Using global statement to update variable is discouraged
    # These tend to be lighweight and confuse pyright
]

exclude = [
    ".bzr",
    ".direnv",
    ".eggs",
    ".git",
    ".hg",
    ".mypy_cache",
    ".nox",
    ".pants.d",
    ".ruff_cache",
    ".svn",
    ".tox",
    ".venv",
    "__pypackages__",
    "_build",
    "buck-out",
    "build",
    "dist",
    "node_modules",
    "venv",
    "docs",
]

# Exclude a variety of commonly ignored directories.
[tool.ruff.lint.per-file-ignores]
"tests/*.py" = [
    "S101",
    "D101",  # Missing docstring in public class
    "D102",  # Missing docstring in public method
    "D103",  # Missing docstring in public function
    "D105",  # Missing docstring in magic method
    "D106",  # Missing docstring in public nested class
    "D107",   # Missing docstring in `__init__`
    "ANN001",
    "ANN201",
    "FBT001",
    "FBT002",
    "D100",
    "PLR2004",
    "PD901", #  X is a bad variable name. (pandas)
    "TCH",
    "N803",
    "C901", # Too complex
    "E501", # Line too long
    "PIE790", # Unnecessary pass statement
    "NPY002", # Replace legacy np.random calls
    "PT011", # pytest.raises is too broad
    "PT012", # pytest.raises block should contain a single simple statement
    "ARG001", # Unused function argument
    "ARG002", # Unused method argument
    "PLR5501", # Use elif instead of else then if
    "SIM102", # Use a single if statement instead of nested if
    "RUF013", # PEP 484 prohibits implicit Optional
    "T201", # print found
    "RET505", # Unnecessary else after return
    "PLR0912", # Too many branches
]
"__init__.py" = ["I002", "D104", "RUF022"] # Missing docstring in public package, __all__ is not sorted
"scripts/*.py" = [
    "D100", # Missing docstring in public module
    "PTH123", # open() should be replaced by Path.open()
    "SIM115", # Use a context manager for opening files
    "PTH118", # os.path.join() should be replaced by Path
    "PTH100", # os.path.abspath() should be replaced by Path.resolve()
    "T201", # print found
    "E501", # Line too long
    "RET505", # Unnecessary else after return
]
"examples/*" = [
    "INP001", "I002", "E741", "D101", "D103", "T20",
    "D415", "ERA001", "E402", "E501", "D100", "NPY002",
    "PD901"
]
"docs/*" = ["INP001"]
# Temporarily ignore docstring and line length issues in existing files
"src/tabpfn_extensions/**/*.py" = [
    "D100", # Missing docstring in public module
    "D101", # Missing docstring in public class
    "D102", # Missing docstring in public method
    "D103", # Missing docstring in public function
    "D105", # Missing docstring in magic method
    "D107", # Missing docstring in public __init__
    "D205", # Missing blank line after summary
    "D415", # First line should end with a period
    "D417", # Missing argument description in docstring
    "E501", # Line too long
    "B904", # Within an except clause, raise exceptions with from err
    "NPY002", # Replace legacy np.random calls
    "TC001", # Move application import into a type-checking block
    "RET505", # Unnecessary else/elif after return statement
    "FBT001", # Boolean-typed positional argument in function definition
    "FBT002", # Boolean default positional argument in function definition
    "ARG001", # Unused function argument
    "ARG002", # Unused method argument
    "C901", # Function is too complex
    "PLR0912", # Too many branches
    "PIE790", # Unnecessary pass statement
    "PD011", # Use .to_numpy() instead of .values
    "RUF022", # __all__ is not sorted
]


[tool.ruff.lint.isort]
known-first-party = ["tabpfn"]
known-third-party = ["sklearn"]
no-lines-before = ["future"]
required-imports = ["from __future__ import annotations"]
combine-as-imports = true
extra-standard-library = ["typing_extensions"]
force-wrap-aliases = true

[tool.ruff.lint.pydocstyle]
convention = "google"

[tool.ruff.lint.pylint]
max-args = 10 # Changed from default of 5

[tool.mypy]
python_version = "3.10"
packages = ["src/tabpfn_extensions", "tests"]

show_error_codes = true

warn_unused_configs = true # warn about unused [tool.mypy] lines

follow_imports = "normal"      # Type check top level api code we use from imports
ignore_missing_imports = true  # Temporarily set to true to handle modules without stubs

disallow_untyped_defs = false  # Temporarily set to false while we add types
disallow_untyped_decorators = false
disallow_incomplete_defs = false

no_implicit_optional = true
check_untyped_defs = true

warn_return_any = true


[[tool.mypy.overrides]]
module = ["tests.*"]
disallow_untyped_defs = false       # Sometimes we just want to ignore verbose types
disallow_untyped_decorators = false # Test decorators are not properly typed
disallow_incomplete_defs = false    # Sometimes we just want to ignore verbose types
disable_error_code = ["var-annotated"]

[[tool.mypy.overrides]]
module = [
]
ignore_errors = true

[[tool.mypy.overrides]]
module = ["sklearn.*"]
ignore_missing_imports = true

[tool.pyright]
include = ["src", "tests"]

pythonVersion = "3.10"
typeCheckingMode = "strict"

strictListInference = true
strictSetInference = true
strictDictionaryInference = false
reportImportCycles = false
reportMissingSuperCall = true
reportMissingTypeArgument = false
reportOverlappingOverload = true
reportIncompatibleVariableOverride = true
reportIncompatibleMethodOverride = true
reportInvalidTypeVarUse = true
reportCallInDefaultInitializer = true
reportImplicitOverride = true
reportUnknownMemberType = false
reportUnknownParameterType = false
reportUnknownVariableType = false
reportUnknownArgumentType = false
reportUnknownLambdaType = false
reportPrivateUsage = false
reportUnnecessaryCast = false
reportUnusedFunction = false
reportMissingTypeStubs = false
reportPrivateImportUsage = false
reportUnnecessaryComparison = false
reportConstantRedefinition = false
reportUntypedFunctionDecorator = false

[tool.changelog-bot.towncrier_changelog]
enabled = true
changelog_skip_label = "no changelog needed"

[tool.towncrier]
directory = "changelog"
filename = "CHANGELOG.md"
start_string = "## [Unreleased]\n"
title_format = "## [{version}] - {project_date}"
issue_format = "[#{issue}](https://github.com/PriorLabs/tabpfn-extensions/pull/{issue})"

[[tool.towncrier.type]]
directory = "breaking"
name = "Breaking Changes"
showcontent = true

[[tool.towncrier.type]]
directory = "added"
name = "Added"
showcontent = true

[[tool.towncrier.type]]
directory = "changed"
name = "Changed"
showcontent = true

[[tool.towncrier.type]]
directory = "fixed"
name = "Fixed"
showcontent = true

[[tool.towncrier.type]]
directory = "deprecated"
name = "Deprecated"
showcontent = true

[[tool.towncrier.section]]
path = ""
name = ""