The Open Source Risk project aims at establishing a transparent peer-reviewed framework for pricing and risk analysis. Open Source Risk Engine (ORE) is based on QuantLib (http://quantlib.org) and consists of three libraries written in C++.
Similar to QuantLib-SWIG, ORE-SWIG intends to provide the means to use the ORE libraries side by side with QuantLib from a number of languages including Python, Ruby, Perl, Java and C#.
For a series of in depth tutorials on installing and/or building python wrappers and wheels, please refer to this document: Tutorials Index. For a general overview on using the ORE python wrapper, continue reading below.
ORE and ORE-SWIG can be downloaded from the OpenSourceRisk project site which points to the Open Source Risk repositories on GitHub.
The ORE-SWIG project directory contains QuantExt-SWIG, OREData-SWIG and OREAnalytics-SWIG folders. Within the ORE-SWIG project directory, pull in the QuantLib-SWIG project by running
git submodule init
git submodule update
Prerequisite for building the wrappers is building the latest ORE release following the steps outlined in ORE's user guide, e.g. using CMake.
To build ORE-SWIG on Windows, macOS or Linux using CMake and Ninja: Edit the top level CMakeLists.txt to select/deselect specific library/language folders, create a subdirectory "build", change to the build directory, and run
cmake \
-G Ninja \
-D ORE=<ORE Root Directory> \
[-D BOOST_ROOT=<Top level boost include directory> ] \
[-D BOOST_LIB=<Location of the compiled boost libraries> ] \
[-D PYTHON_LIBRARY=<Full name including path to the 'libpython*' library> ] \
[-D PYTHON_INCLUDE_DIR=<Directory that contains Python.h> ] \
..
ninja
To build on Windows using CMake and an existing Visual Studio installation you can e.g. run this from the top-level oreswig directory
mkdir builddir
cmake -G "Visual Studio 17 2022" \
-A x64 \
-D SWIG_DIR=C:\dev\swigwin\Lib \
-D SWIG_EXECUTABLE=C:\dev\swigwin\swig.exe \
-D ORE=<ORE Root Directory> \
-D BOOST_INCLUDEDIR=<Top level boost include directory> \
-D BOOST_LIBRARYDIR=<Location of the compiled boost libraries>\
-D ORE_USE_ZLIB=[ON/OFF] (depending on your ORE build)
-D MSVC_LINK_DYNAMIC_RUNTIME=[ON/OFF] (depending on your ORE build)
-D Boost_NO_SYSTEM_PATHS=ON
-S OREAnalytics-SWIG/Python \
-B builddir
cmake --build builddir -v
To try e.g. an OREAnalytics Python example, update your PYTHONPATH so that it includes the directory that contains the newly built python module and associated native library (both in ORE-SWIG/build/OREAnalytics-SWIG/Python), change to ORE-SWIG/OREAnalytics-SWIG/Python/Examples and run
python ore.py
You can also try the IPython example in the same directory: Launch
jupyter notebook
wait for your browser to open, select ore.ipy from the list of files and then run all cells.
Also note the test suite in ORE-SWIG/OREAnalytics-SWIG/Python/test:
python OREAnalyticsTestSuite.py
When the QuantExt-SWIG and OREData-SWIG modules are built as well (edit top-level CMakeLists.txt) then similar test suites in ORE-SWIG/OREData-SWIG/Python/test and ORE-SWIG/QuantExt-SWIG/Python/test can be run
python OREDataTestSuite.py
python QuantExtTestSuite.py
To try a simple OREAnalytics Java example, change to ORE-SWIG/OREAnalytics-SWIG/Java/Examples and run
java -Djava.library.path=../../../build/OREAnalytics-SWIG/Java \
-jar ../../../build/OREAnalytics-SWIG/Java/ORERunner.jar \
Input/ore.xml
On Windows you can also use the following steps to build the Python bindings using the provided setup.py script:
1. Include SWIG path to the Path environment variable, e.g.
set Path=%Path%;C:\swigwin-3.0.12
2. Add PYTHON_INCLUDE and PYTHON_LIB variables to the system environment, e.g.
set PYTHON_INCLUDE="C:\Users\Name\AppData\Local\Continuum\anaconda3\include"
set PYTHON_LIB="C:\Users\Name\AppData\Local\Continuum\anaconda3\libs"
3. Add BOOST_ROOT and BOOST_LIB variables to the system environment, e.g.
set BOOST_ROOT=C:\repos\boost_1_65_1
set BOOST_LIB=C:\repos\boost_1_65_1\lib\x64\lib\lib
4. Add ORE_DIR and QL_DIR variables to the system environment, e.g.
set ORE_DIR=C:\dev\ORE
5. Change to directory OREAnalytics-SWIG/Python and run the following
python scripts to build and install ORE Analytics Python module:
cd Python
python setup.py wrap
python setup.py build
python setup.py install
6. Try examples (all work):
cd Examples
python ore.py
python swap.py
python market.py
python commodityforward.py
python conventions.py
python portfolio.py
ORE-SWIG initially contains the framework for building the wrappers including QuantLib wrappers and a limited number of wrapped classes in each of the three ORE libraries to demonstrate the principle. Overall, ORE contains more than 500 classes that could be added to ORE-SWIG over time.
The easiest way to contribute additional interface files is through pull requests on GitHub. Get a GitHub account if you don't have it already and clone the repository at https://github.com/OpenSourceRisk/ORE-SWIG with the "Fork" button in the top right corner of the page. Check out your clone to your machine, code away, push your changes to your clone and submit a pull request; instructions are available at https://help.github.com/articles/fork-a-repo. (In case you need them, more detailed instructions for creating pull requests are at https://help.github.com/articles/using-pull-requests, and a basic guide to GitHub is at https://guides.github.com/activities/hello-world/.
We're looking forward to your contributions.