add C++ components

cpp/include/segmentation.hpp

@@ -0,0 +1,11 @@
+#pragma once
+include "xtensor/"
+
+namespace libneutorch{
+
+template <class T, xt::layout_type L>
+auto label2affinity(pytensor<T, 3, L> label){
+
+}
+
+} // namespace libneutorch

cpp/main.cpp

@@ -0,0 +1,20 @@
+#include <pybind11/pybind11.h>
+
+
+PYBIND11_MODULE(libneutorch, m) {
+    m.doc() = R"pbdoc(
+        libneutorch
+        -----------------------
+        .. currentmodule:: libneutorch
+        .. autosummary::
+           :toctree: _generate
+           warp3d
+    )pbdoc";
+
+    m.def("warp3d", &warp3d, R"pbdoc(
+        Warp 3d image 
+
+        used for patch augmentation.
+    )pbdoc");
+
+}

neutorch/__version__.py

@@ -0,0 +1 @@
+__version__ = "0.0.2"

neutorch/dataset/patch.py

@@ -19,6 +19,7 @@ def __init__(self, image: np.ndarray, label: np.ndarray,
         assert image.shape == label.shape
         self.image = image
         self.label = label
+        self.target = None
         self.delayed_shrink_size = delayed_shrink_size
 
     def accumulate_delayed_shrink_size(self, shrink_size: tuple):
@@ -61,4 +62,17 @@ def shape(self):
     @property
     @lru_cache
     def center(self):
-        return tuple(ps // 2 for ps in self.shape)
+        return tuple(ps // 2 for ps in self.shape)
+
+    @property
+    def target(self):
+        if self.target is None:
+            assert np.issubdtype(self.label.dtype, np.floating)
+            return self.label
+        else:
+            return self.target
+
+    @property
+    @lru_cache
+    def affinity_map(self):
+

neutorch/dataset/transform.py

@@ -536,3 +536,7 @@ def transform(self, patch: Patch):
                 strength=random.randint(1, self.max_strength),
                 radius = (patch.shape[-1] + patch.shape[-2]) // 4,
             )
+
+class LabelAsTarget()
+
+class Label2Affinity

requirements.txt

@@ -3,5 +3,6 @@ torchvision
 toml
 h5py
 tensorboard
+pybind11
+scikit-image
 opencv-python
-scikit-image

setup.py

@@ -1,15 +1,56 @@
 #!/usr/bin/env python
+import os
+import re
 
 from setuptools import setup, find_packages
 
+from pybind11.setup_helpers import Pybind11Extension, build_ext
+
+
+PACKAGE_DIR = os.path.dirname(os.path.abspath(__file__))
+
+with open(os.path.join(PACKAGE_DIR, 'requirements.txt')) as f:
+    requirements = f.read().splitlines()
+    requirements = [l for l in requirements if not l.startswith('#')]
+
+with open("README.md", "r") as fh:
+    long_description = fh.read()
+
+VERSIONFILE = os.path.join(PACKAGE_DIR, "neutorch/__version__.py")
+verstrline = open(VERSIONFILE, "rt").read()
+VSRE = r"^__version__ = ['\"]([^'\"]*)['\"]"
+mo = re.search(VSRE, verstrline, re.M)
+if mo:
+    version = mo.group(1)
+else:
+    raise RuntimeError("Unable to find version string in %s." %
+                       (VERSIONFILE, ))
+
+
+ext_modules = [
+    Pybind11Extension("libneutorch",
+        ["cpp/main.cpp"],
+        # Example: passing in the version to the compiled code
+        define_macros = [('VERSION_INFO', version)],
+        ),
+]
+
 setup(
     name='neutorch',
-    version='0.0.1',
+    version=version,
     description='Deep Learning for brain connectomics using PyTorch',
+    long_description=long_description,
+    long_description_content_type="text/markdown",
     author='Jingpeng Wu',
     author_email='[email protected]',
     url='https://github.com/brain-connectome/neutorch',
     packages=find_packages(exclude=['bin']),
+    cmdclass={"build_ext": build_ext},
+    ext_modules=ext_modules,
+    install_requires=requirements,
+    tests_require=[
+        'pytest',
+    ],
     entry_points='''
         [console_scripts]
         neutrain-tbar=neutorch.cli.train_tbar:train
@@ -25,4 +66,6 @@
         "Operating System :: OS Independent",
         "Programming Language :: Python :: 3",
     ],
+    python_requires='>=3',
+    zip_safe=False,
 )