The instructions below assume that ELL was obtained from GitHub using git. The git client is typically installed by default on Ubuntu systems, but if it isn't, open a terminal and type
sudo apt-get install git
In a terminal, clone the ELL repository by typing
git clone https://github.com/Microsoft/ELL.git
We recommend using the apt-get package manager to download and install prerequisites. First, make sure that apt-get is up to date, by typing
sudo apt-get -y update
Next, you will need gcc and CMake. They are often installed by default on Ubuntu systems, but to confirm type
sudo apt-get install -y gcc cmake
ELL requires gcc 5 and above for C++14 support.
You will also need the dev version of LLVM-3.9. At the time of writing this document, apt-get doesn't yet have the required version of LLVM. To check this, type
apt-cache show llvm-dev
and look for the version number. If the version number is 3.9 or greater, you can simply type sudo apt-get install -y llvm-dev. Otherwise, do the following:
sudo apt-get install -y wget
wget -O - http://apt.llvm.org/llvm-snapshot.gpg.key | sudo apt-key add -
sudo apt-add-repository "deb http://apt.llvm.org/xenial/ llvm-toolchain-xenial-3.9 main"
sudo apt-get -y update
sudo apt-get install -y llvm-3.9-dev
Install the BSD editline and history libraries
sudo apt-get install -y libedit-dev
sudo apt-get install zlibc zlib1g zlib1g-dev
The language bindings for Python are generated using SWIG. To build them, you need to install SWIG version 3.0.12 or later.
wget http://prdownloads.sourceforge.net/swig/swig-3.0.12.tar.gz
tar zxvf swig-3.0.12.tar.gz && cd swig-3.0.12
./configure --without-pcre && make && sudo make install
The tutorials require Python 3.6, we recommend you use Miniconda, which works well with Jupyter notebooks and provides a way to manage different Python versions. Note: you can also use the full Anaconda if you already have that installed.
If you build ELL from a conda Python 3.6 environment then the Python language bindings will be built and you can run the
tutorials.
To configure the Python 3.6 environment using Miniconda:
# Create the environment
conda create -n py36 anaconda python=3
# Activate the environment
source activate py36
Now you have an activated conda Python 3.6 environment, and if you build ELL from this environment then the Python language bindings will be built and you can run the tutorials.
ELL can optionally take advantage of the optimized linear algebra libraries in OpenBLAS, and generate code documentation using Doxygen. To install these optional dependencies, type
sudo apt-get install -y libopenblas-dev doxygen
Next, use CMake to create a makefile. In the repository root directory, create a build subdirectory and change to that directory:
mkdir build
cd build
Then, invoke cmake:
cmake ..
Important: don't forget the two dots (..) at the end of the command!
Now you can build ELL by typing:
make
and lastly, so you can run the tutorials, the following will build the Python language bindings for ELL:
make _ELL_python
The generated executables will appear in a directory named ELL/build/bin.
The create code documentation, type
make doc
The Doxygen generated files will appear in ELL/build/doc.