quantconnect--lean/Algorithm.Python

QuantConnect Python Algorithm Project:

Before we enable python support, follow the installation instructions to get LEAN running C# algorithms in your machine.

Install Python 3.6:

Windows

1. Use the Windows x86-64 MSI Python 3.6.6 installer from python.org or Anaconda for Windows installer.
2. When asked to select the features to be installed, make sure you select "Add python.exe to Path"
3. [Optional] Create PYTHONHOME system variables which value must be the location of your python installation (e.g. C:\Python36amd64 or C:\Anaconda3):
   1. Right mouse button on My Computer. Click Properties.
   2. Click Advanced System Settings -> Environment Variables -> System Variables
   3. Click New.
      • Name of the variable: PYTHONHOME.
      • Value of the variable: python installation path.
4. Install pandas and its dependencies.
5. Reboot computer to ensure changes are propogated.

macOS

1. Use the macOS x86-64 package installer from Anaconda and follow "Installing on macOS" instructions from Anaconda documentation page.
2. Install pandas and its dependencies.

Note: If you encounter the "System.DllNotFoundException: python3.6m" runtime error when running Python algorithms on macOS:

1. Find libpython3.6m.dylib in your Python installation folder. If you installed Python with Anaconda, it may be find at

/Users/{your_user_name}/anaconda3/lib/libpython3.6m.dylib

2. Open Lean/Launcher/bin/Debug/Python.Runtime.dll.config, add the following text and save:

 <configuration>
    <dllmap dll="python3.6m" target="{the path in step 1 including libpython3.6m.dylib}" os="!windows"/>
</configuration>

Note: Specify the install of v3.6.6 exactly, i.e. if with conda conda install python=3.6.6 as this is a known compatible version and other versions may have issues as of this writing.

Linux

By default, miniconda is installed in the users home directory ($HOME):

export PATH="$HOME/miniconda3/bin:$PATH"
wget https://cdn.quantconnect.com/miniconda/Miniconda3-4.3.31-Linux-x86_64.sh
bash Miniconda3-4.3.31-Linux-x86_64.sh -b
rm -rf Miniconda3-4.3.31-Linux-x86_64.sh
sudo ln -s $HOME/miniconda3/lib/libpython3.6m.so /usr/lib/libpython3.6m.so
conda update -y python conda pip
conda install -y cython pandas

Note: There is a known issue with python 3.6.5 that prevents pythonnet installation, please upgrade python to version 3.6.6:

conda install -y python=3.6.6

Run python algorithm

1. Update the config to run the python algorithm:

"algorithm-type-name": "BasicTemplateAlgorithm",
"algorithm-language": "Python",
"algorithm-location": "../../../Algorithm.Python/BasicTemplateAlgorithm.py",

2. Rebuild LEAN.
3. Run LEAN. You should see the same result of the C# algorithm you tested earlier.

---

Python.Runtime.dll compilation

LEAN users do not need to compile Python.Runtime.dll. The information below is targeted to developers who wish to improve it.

Download QuantConnect/pythonnet github clone or downloading the zip. If downloading the zip - unzip to a local pathway.

Note: QuantConnect's version of pythonnet is an enhanced version of pythonnet with added support for System.Decimal and System.DateTime.

Below we can find the compilation flags that create a suitable Python.Runtime.dll for each operating system.

Windows

msbuild pythonnet.sln /nologo /v:quiet /t:Clean;Rebuild /p:Platform=x64 /p:PythonInteropFile="interop36.cs" /p:Configuration=ReleaseWin /p:DefineConstants="PYTHON36,PYTHON3,UCS2"

macOS

msbuild pythonnet.sln /nologo /v:quiet /t:Clean;Rebuild /p:Platform=x64 /p:PythonInteropFile="interop36m.cs" /p:Configuration=ReleaseMono /p:DefineConstants="PYTHON36,PYTHON3,UCS4,MONO_OSX,PYTHON_WITH_PYMALLOC"

Linux

msbuild pythonnet.sln /nologo /v:quiet /t:Clean;Rebuild /p:Platform=x64 /p:PythonInteropFile="interop36m.cs" /p:Configuration=ReleaseMono /p:DefineConstants="PYTHON36,PYTHON3,UCS4,MONO_LINUX,PYTHON_WITH_PYMALLOC"