install.md

Installation steps

The procedure was tested on Ubuntu 18.04 with an installed CUDA driver of version 10.1.

Environmental variables:

• $PD_MESH_NET_ROOT: Root of this folder.
• $PYG_ROOT: Root of PyTorch Geometric.

Steps

It will be assumed that virtualenvwrapper is installed. Alternatively, the creation and activation of the virtualenv (mkvirtualenv and workon in the instructions below) can be performed according to the official PyPA instructions.

Create a virtualenv:

mkvirtualenv pd_mesh_net --python=$(which python3)

Install PyTorch Geometric:

export THIRD_PARTY_ROOT=$PD_MESH_NET_ROOT/third_party
export PYG_ROOT=$THIRD_PARTY_ROOT/pytorch_geometric

• Clone PyTorch Geometric:

  cd $THIRD_PARTY_ROOT
  git clone -n https://github.com/rusty1s/pytorch_geometric.git
  cd $PYG_ROOT
  git checkout 082fb83

• If not already in ~/.bashrc, add CUDA to $PATH and $CPATH:

  export PATH=/usr/local/cuda-10.1/bin${PATH:+:${PATH}}
  export CPATH=/usr/local/cuda-10.1/include:$CPATH

• Add CUDA to $LD_LIBRARY_PATH:

  export LD_LIBRARY_PATH=/usr/local/cuda-10.1/lib64:$LD_LIBRARY_PATH

• Install dependencies (in virtualenv, i.e., run workon pd_mesh_net first):

  pip install torch==1.4.0

  • Installed torch-1.4.0 (https://files.pythonhosted.org/packages/24/19/4804aea17cd136f1705a5e98a00618cb8f6ccc375ad8bfa437408e09d058/torch-1.4.0-cp36-cp36m-manylinux1_x86_64.whl)

  pip install numpy

  • Installed numpy-1.18.1 (https://files.pythonhosted.org/packages/62/20/4d43e141b5bc426ba38274933ef8e76e85c7adea2c321ecf9ebf7421cedf/numpy-1.18.1-cp36-cp36m-manylinux1_x86_64.whl)

  pip install --verbose --no-cache-dir torch-scatter==1.4.0

  • Installed torch-scatter-1.4.0 (https://files.pythonhosted.org/packages/b8/c3/8bad887ffa55c86f120ef5ae252dc0e357b3bd956d9fbf45242bacc46290/torch_scatter-1.4.0.tar.gz)

  pip install --verbose --no-cache-dir torch-sparse==0.4.3

  • Installed scipy-1.4.1 (https://files.pythonhosted.org/packages/dc/29/162476fd44203116e7980cfbd9352eef9db37c49445d1fec35509022f6aa/scipy-1.4.1-cp36-cp36m-manylinux1_x86_64.whl)
  • Installed torch-sparse-0.4.3 (https://files.pythonhosted.org/packages/08/4e/a268613fa6a92ffbc65b89e66fc8be5590801937185007f0f7bcb75ea21f/torch_sparse-0.4.3.tar.gz)

  pip install --verbose --no-cache-dir torch-cluster==1.4.5

  • Installed torch_cluster-1.4.5 (https://files.pythonhosted.org/packages/c3/70/1d827d6fd1e03bb5ae84852dd0070c6574105c37e7b935284f6e990932db/torch_cluster-1.4.5.tar.gz)

  pip install --verbose --no-cache-dir torch-spline-conv==1.1.1

  • Installed torch_spline_conv-1.1.1 (https://files.pythonhosted.org/packages/5e/77/5420584cdb1514c580722ca4bc482a509105d64b7c70246e9dc4a3e6d3c5/torch_spline_conv-1.1.1.tar.gz)

• Install PyTorch Geometric:

  cd $PYG_ROOT

  • Comment lines 30 and 31 in $PYG_ROOT/setup.py (i.e., lines starting for url and download_url).

  workon pd_mesh_net
  python setup.py install

• Install optional PyTorch-Geometric dependencies:

  pip install tensorboard

  As a reference, the above command installed the following package versions:

  • Installed absl-py-0.9.0 (https://files.pythonhosted.org/packages/1a/53/9243c600e047bd4c3df9e69cfabc1e8004a82cac2e0c484580a78a94ba2a/absl-py-0.9.0.tar.gz)
  • Installed cachetools-4.0.0 (https://files.pythonhosted.org/packages/08/6a/abf83cb951617793fd49c98cb9456860f5df66ff89883c8660aa0672d425/cachetools-4.0.0-py3-none-any.whl)
  • Installed google-auth-1.11.0 (https://files.pythonhosted.org/packages/1c/6d/7aae38a9022f982cf8167775c7fc299f203417b698c27080ce09060bba07/google_auth-1.11.0-py2.py3-none-any.whl)
  • Installed google-auth-oauthlib-0.4.1 (https://files.pythonhosted.org/packages/7b/b8/88def36e74bee9fce511c9519571f4e485e890093ab7442284f4ffaef60b/google_auth_oauthlib-0.4.1-py2.py3-none-any.whl)
  • Installed grpcio-1.26.0 (https://files.pythonhosted.org/packages/17/8f/f79c5c174bebece41f824dd7b1ba98da45dc2d4c373b38ac6a7f6a5acb5e/grpcio-1.26.0-cp36-cp36m-manylinux2010_x86_64.whl)
  • Installed markdown-3.1.1 (https://files.pythonhosted.org/packages/c0/4e/fd492e91abdc2d2fcb70ef453064d980688762079397f779758e055f6575/Markdown-3.1.1-py2.py3-none-any.whl)
  • Installed oauthlib-3.1.0 (https://files.pythonhosted.org/packages/05/57/ce2e7a8fa7c0afb54a0581b14a65b56e62b5759dbc98e80627142b8a3704/oauthlib-3.1.0-py2.py3-none-any.whl)
  • Installed protobuf-3.11.2 (https://files.pythonhosted.org/packages/ca/ac/838c8c8a5f33a58132dd2ad2a30329f6ae1614a9f56ffb79eaaf71a9d156/protobuf-3.11.2-cp36-cp36m-manylinux1_x86_64.whl)
  • Installed pyasn1-0.4.8 (https://files.pythonhosted.org/packages/62/1e/a94a8d635fa3ce4cfc7f506003548d0a2447ae76fd5ca53932970fe3053f/pyasn1-0.4.8-py2.py3-none-any.whl)
  • Installed pyasn1-modules-0.2.8 (https://files.pythonhosted.org/packages/95/de/214830a981892a3e286c3794f41ae67a4495df1108c3da8a9f62159b9a9d/pyasn1_modules-0.2.8-py2.py3-none-any.whl)
  • Installed requests-oauthlib-1.3.0 (https://files.pythonhosted.org/packages/a3/12/b92740d845ab62ea4edf04d2f4164d82532b5a0b03836d4d4e71c6f3d379/requests_oauthlib-1.3.0-py2.py3-none-any.whl)
  • Installed rsa-4.0 (https://files.pythonhosted.org/packages/02/e5/38518af393f7c214357079ce67a317307936896e961e35450b70fad2a9cf/rsa-4.0-py2.py3-none-any.whl)
  • Installed tensorboard-2.1.0 (https://files.pythonhosted.org/packages/40/23/53ffe290341cd0855d595b0a2e7485932f473798af173bbe3a584b99bb06/tensorboard-2.1.0-py3-none-any.whl)
  • Installed werkzeug-0.16.0 (https://files.pythonhosted.org/packages/ce/42/3aeda98f96e85fd26180534d36570e4d18108d62ae36f87694b476b83d6f/Werkzeug-0.16.0-py2.py3-none-any.whl)

Install PyMesh

• Make sure your version of CMake is >= 3.11 (follow, e.g., https://answers.ros.org/question/293119/how-can-i-updateremove-cmake-without-partially-deleting-my-ros-distribution/?answer=297523#post-id-297523 if this is not the case).
• Install PyMesh from the Github repo:

  workon pd_mesh_net
  pip install git+git://github.com/PyMesh/PyMesh.git@93d182c0a7cee446e89fac74033347d900054af4

Install other dependencies

pip install pyyaml

As a reference, the above command installed the following package versions:

• Installed pyyaml-5.3

Install package

cd $PD_MESH_NET_ROOT
workon pd_mesh_net
pip install -e .