Skip to content

Latest commit

 

History

History
257 lines (170 loc) · 8.38 KB

INSTALL.md

File metadata and controls

257 lines (170 loc) · 8.38 KB
Raw

Linux-based OS

Easy-install procedure on Ubuntu 18/19/20, and Debian 8/9

Jiminy

Dependencies installation

There is not requirement to install jiminy_py on linux if one does not want to build it. Nevertheless, this package does not provide the backend viewer gepetto-gui (still, the backend meshcat is available).

(optional) Gepetto viewer

The first step to install gepetto-gui is to setup the APT repository robotpkg to have access to compiled binaries.

sh -c "echo 'deb [arch=amd64] http://robotpkg.openrobots.org/packages/debian/pub bionic robotpkg' >> /etc/apt/sources.list.d/robotpkg.list" && \
curl http://robotpkg.openrobots.org/packages/debian/robotpkg.key | sudo apt-key add -
apt update

Once done, it is straightforward to install the required package for Python 3.6.

For Python 3.6

sudo apt install -y robotpkg-gepetto-viewer=4.4.0 robotpkg-py36-qt4-gepetto-viewer-corba=5.1.2 robotpkg-py36-omniorbpy

Install Jiminy Python package

The project is available on PyPi and therefore can be install easily using pip.

python -m pip install jiminy-py

Gym Jiminy (Python 3 only)

Dependencies installation

Tensorflow>=2.0 and Pytorch>=1.13 with GPU support dependencies (Cuda 10.1 and CuDNN 7.6)

Amazing tutorial for Ubuntu 18 to install Tensorflow, along with CUDA toolkit: https://medium.com/better-programming/install-tensorflow-1-13-on-ubuntu-18-04-with-gpu-support-239b36d29070

Once done, Pytorch can be installed following the official "getting started" instructions: https://pytorch.org/get-started/locally/

(optional) stable_baselines3

Installing the Python packages stable_baselines3==0.9 is required to run some of the provided examples, though it is not required to use gym_jiminy.

python -m pip install stable-baselines3[extra]
(optional) tianshou

Installing the Python packages tianshou==0.3.0 is required to run some of the provided examples, though it is not required to use gym_jiminy.

python -m pip install tianshou
(optional) ray[rllib]

Installing the Python packages ray==1.0.0 are required to run some of the provided examples, though it is not required to use gym_jiminy. It can be easily installed using pip for any OS and Python 3.6/3.7/3.8. The installation instructions are available here.

Install Gym Jiminy learning Python package

python -m pip install gym-jiminy

Build Jiminy from source on Ubuntu 18+ (excluding dependencies)

First, one must install the pre-compiled libraries of the dependencies. Most of them are available on robotpkg APT repository. Just run the bash script to install them automatically for Ubuntu 18 and upward. It should be straightforward to adapt it to any other distribution for which robotpkg is available.

sudo ./build_tools/easy_install_deps_ubuntu.sh

You are now ready to build and install Jiminy itself.

RootDir=".... The location of jiminy repository ...."

BUILD_TYPE="Release"
InstallDir="$RootDir/install"

mkdir "$RootDir/build" "$InstallDir"
cd "$RootDir/build"
cmake "$RootDir" -DCMAKE_INSTALL_PREFIX="$InstallDir" \
      -DBoost_NO_SYSTEM_PATHS=OFF \
      -DBUILD_TESTING=ON -DBUILD_EXAMPLES=ON -DBUILD_PYTHON_INTERFACE=ON \
      -DCMAKE_BUILD_TYPE="$BUILD_TYPE"
make install -j2

Building from source (including dependencies)

Prerequisites

sudo apt install -y gnupg curl wget build-essential cmake doxygen graphviz
python -m pip install "numpy<1.20"

Jiminy dependencies build and install

Just run the bash script already available.

BUILD_TYPE="Release" ./build_tools/build_install_deps_linux.sh

Build Procedure

RootDir=".... The location of jiminy repository ...."
PythonVer=".... Your version X.Y of Python, for instance 3.8 ...."

BUILD_TYPE="Release"
InstallDir="$RootDir/install"

unset Boost_ROOT

mkdir "$RootDir/build"
cd "$RootDir/build"
cmake "$RootDir" -DCMAKE_INSTALL_PREFIX="$InstallDir" -DCMAKE_PREFIX_PATH="$InstallDir" \
      -DBOOST_ROOT="$InstallDir" -DBoost_INCLUDE_DIR="$InstallDir/include" \
      -DBoost_NO_SYSTEM_PATHS=TRUE -DBoost_NO_BOOST_CMAKE=TRUE \
      -DBoost_USE_STATIC_LIBS=OFF -DPYTHON_REQUIRED_VERSION="$PythonVer" \
      -DBUILD_TESTING=ON -DBUILD_EXAMPLES=ON -DBUILD_PYTHON_INTERFACE=ON \
      -DCMAKE_CXX_FLAGS="-DURDFDOM_STATIC" -DCMAKE_BUILD_TYPE="$BUILD_TYPE"
make install -j2

mkdir -p "$HOME/.local/lib/python${PythonVer}/site-packages"
echo "$InstallDir/lib/python${PythonVer}/site-packages" \
> "$HOME/.local/lib/python${PythonVer}/site-packages/user_site.pth"

Windows OS

Easy-install procedure on Windows (Python 3 only)

Jiminy

Dependencies installation

Install python3 3.6/3.7/3.8 (available on Microsoft store), and Microsoft Visual C++ Redistributable for Visual Studio 2015, 2017 and 2019.

Fixing Meshcat viewer

Installing the master branch of meshcat from github instead of the latest official release on Pypi should do the trick.

python -m pip install --upgrade git+https://github.com/rdeits/meshcat-python.git@master

Install Jiminy Python package

The project is available on PyPi and therefore can be install easily using pip.

python -m pip install jiminy-py

Gym Jiminy

Dependencies installation

Tensorflow>=2.0 and Pytorch>=1.13 with GPU support dependencies (Cuda 10.1 and CuDNN 7.6)

See this tutorial: https://towardsdatascience.com/installing-tensorflow-with-cuda-cudnn-and-gpu-support-on-windows-10-60693e46e781

Once done, Pytorch can be installed following the official "getting started" instructions: https://pytorch.org/get-started/locally/

(optional) stable_baselines3 / ray[rllib] / tianshou

Installing the Python packages stable_baselines3, tianshou, ray are required to run all the provided examples, though they are not required to use gym_jiminy. They can easily be installed using pip. Pick the one you prefer!

python -m pip install stable-baselines3[extra]
python -m pip install tianshou
python -m pip install ray[rllib]

Install Gym Jiminy learning Python package

python -m pip install gym-jiminy

Building from source (including dependencies)

Prerequisites

You have to preinstall by yourself the (free) MSVC 2019 toolchain.

Then, install numpy and wheel.

python -m pip install wheel "numpy<1.20"

Jiminy dependencies build and install

Now you can simply run the powershell script already available.

$env:BUILD_TYPE = "Release"
& './build_tools/build_install_deps_windows.ps1'

Build Procedure

You are finally ready to build Jiminy itself.

$RootDir = ".... The location of jiminy repository ...."

$env:BUILD_TYPE = "Release"
$RootDir = $RootDir -replace '\\', '/'
$InstallDir = "$RootDir/install"

if (Test-Path env:Boost_ROOT) {
  Remove-Item env:Boost_ROOT
}

if (-not (Test-Path -PathType Container $RootDir/build)) {
  New-Item -ItemType "directory" -Force -Path "$RootDir/build"
}
Set-Location -Path $RootDir/build
cmake "$RootDir" -G "Visual Studio 16 2019" -T "v142" -DCMAKE_GENERATOR_PLATFORM=x64 `
      -DCMAKE_INSTALL_PREFIX="$InstallDir" -DCMAKE_MODULE_PATH="$InstallDir" `
      -DBOOST_ROOT="$InstallDir" -DBoost_INCLUDE_DIR="$InstallDir/include" `
      -DBoost_NO_SYSTEM_PATHS=TRUE -DBoost_NO_BOOST_CMAKE=TRUE `
      -DBoost_USE_STATIC_LIBS=OFF `
      -DBUILD_TESTING=ON -DBUILD_EXAMPLES=ON -DBUILD_PYTHON_INTERFACE=ON `
      -DCMAKE_CXX_FLAGS="/EHsc /bigobj -D_USE_MATH_DEFINES -DBOOST_ALL_NO_LIB -DBOOST_LIB_DIAGNOSTIC -DURDFDOM_STATIC"
cmake --build . --target all --config "${env:BUILD_TYPE}" --parallel 8

if (-not (Test-Path -PathType Container "$RootDir/build/PyPi/jiminy_py/src/jiminy_py/core")) {
  New-Item -ItemType "directory" -Force -Path "$RootDir/build/PyPi/jiminy_py/src/jiminy_py/core"
}
Copy-Item -Path "$InstallDir/lib/boost_numpy*.dll" `
          -Destination "$RootDir/build/PyPi/jiminy_py/src/jiminy_py/core"
Copy-Item -Path "$InstallDir/lib/boost_python*.dll" `
          -Destination "$RootDir/build/PyPi/jiminy_py/src/jiminy_py/core"
Copy-Item -Path "$InstallDir/lib/site-packages/*" `
          -Destination "$RootDir/build/PyPi/jiminy_py/src/jiminy_py" -Recurse

cmake --build . --target install --config "${env:BUILD_TYPE}"