pytorch_connectomics/setup.py at master · PytorchConnectomics/pytorch_connectomics · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
import os
import sys
from pathlib import Path
from distutils.sysconfig import get_python_inc
from setuptools import setup, Extension, find_packages

# Core dependencies - always required
requirements = [
    # Deep learning frameworks
    "torch>=1.8.0",
    "torchvision>=0.9.0",
    "numpy>=1.23.0",  # >=1.23 for matplotlib; mahotas 1.4.18+ compatible with numpy 2.x
    # PyTorch Lightning & MONAI (PRIMARY frameworks)
    "pytorch-lightning>=2.0.0",
    "monai>=0.9.1",
    "torchmetrics>=0.11.0",
    # Configuration management (Hydra/OmegaConf)
    "omegaconf>=2.1.0",
    # Scientific computing
    "scipy>=1.5",
    "scikit-learn>=0.23.1",
    "scikit-image>=0.17.2",
    # Image processing & I/O
    "opencv-python>=4.3.0",
    "h5py>=2.10.0",
    "imageio>=2.9.0",
    # Visualization & logging
    "matplotlib>=3.3.0",
    "tensorboard>=2.2.2",
    # Utilities
    "tqdm>=4.58.0",
    "einops>=0.3.0",
    "psutil>=5.8.0",
    "filelock>=3.0.0",  # cross-platform lock for concurrent chunked HDF5 writes
    # Post-processing (required for segmentation)
    "connected-components-3d>=3.0.0",  # imports as 'cc3d'
    "fastremap>=1.10.0",  # Fast remapping for segmentation labels
    "kimimaro>=1.0.0",  # Skeletonization library
    "crackle-codec>=0.1.0",  # Required by kimimaro for compression
    "mahotas>=1.4.0",  # Image processing (morphological operations, connected components)
    "fastmorph>=1.0.0",  # Multi-label morphological opening (shape_smooth decoder)
    # Build tools
    "Cython>=0.29.22",
]

# Optional dependencies for specific features
extras_require = {
    # Full installation with all recommended features
    "full": [
        "gputil>=1.4.0",
        "nd2>=0.7.0",
        "tifffile>=2021.11.2",
        "wandb>=0.13.0",
        "optuna>=2.10.0",
        "neuroglancer>=1.0.0",
    ],
    # Advanced metrics (skeleton-based)
    "metrics": [
        # Install manually: pip install git+https://github.com/funkelab/funlib.evaluate.git
    ],
    # Development and testing
    "dev": [
        "pytest>=6.0.0",
        "pytest-benchmark>=3.4.0",
    ],
    # Documentation build
    "docs": [
        "sphinx>=4.0",  # Support modern Sphinx versions (including 8.x)
        "pytorch-sphinx-theme>=0.0.19",
        "sphinxcontrib-katex",
        "jinja2>=3.0",
    ],
    # Command-line tools and utilities
    "cli": [
        # Note: just is not available via pip, install separately:
        # - Rust: cargo install just
        # - Homebrew: brew install just
        # - Conda: conda install -c conda-forge just
        # - Arch: pacman -S just
        # - Ubuntu/Debian: apt install just
    ],
    # MedNeXt models (external package)
    # Install separately: pip install -e /projects/weilab/weidf/lib/MedNeXt
    # Or from your local MedNeXt installation path
    "mednext": [
        # Placeholder - install manually from /projects/weilab/weidf/lib/MedNeXt
    ],
    # Mutex Watershed decoder (affinity -> instances, seedless).
    # Backs the 'decode_mutex_watershed' decoder. affogato is a C++/xtensor
    # package published on conda-forge only (the PyPI 'affogato' is an
    # unrelated project). Install manually:
    #     conda install -c conda-forge affogato
    "mws": [
        # Placeholder - install manually: conda install -c conda-forge affogato
    ],
}


def getInclude():
    dirName = get_python_inc()
    return [dirName, os.path.dirname(dirName)]


def setup_package():
    __version__ = "2.0.0"
    url = "https://github.com/zudi-lin/pytorch_connectomics"
    readme_path = Path(__file__).resolve().parent / "README.md"

    setup(
        name="connectomics",
        description="Semantic and instance segmentation toolbox for EM connectomics",
        long_description=readme_path.read_text(encoding="utf-8"),
        long_description_content_type="text/markdown",
        version=__version__,
        url=url,
        license="MIT",
        author="PyTorch Connectomics Contributors",
        python_requires=">=3.8,<3.13",  # Python 3.13 has limited pre-built wheel support
        install_requires=requirements,
        extras_require=extras_require,
        include_dirs=getInclude(),
        packages=find_packages(),
        classifiers=[
            "Development Status :: 4 - Beta",
            "Intended Audience :: Science/Research",
            "License :: OSI Approved :: MIT License",
            "Programming Language :: Python :: 3",
            "Programming Language :: Python :: 3.8",
            "Programming Language :: Python :: 3.9",
            "Programming Language :: Python :: 3.10",
            "Topic :: Scientific/Engineering :: Artificial Intelligence",
            "Topic :: Scientific/Engineering :: Image Processing",
        ],
    )


if __name__ == "__main__":
    # pip install --editable .
    setup_package()