QuantLib-SWIG Language Bindings

QuantLib-SWIG provides language bindings for the QuantLib project, which is a comprehensive open-source software framework for quantitative finance (modeling, trading, and risk management).

Currently, QuantLib-SWIG supports the following languages:

• Python
• C#
• Java
• Scala
• R

The R interface to QuantLib is created using SWIG (Simplified Wrapper and Interface Generator).

Requirements:

• SWIG version 1.3.32 is required.
• Note: SWIG 1.3.31 generates QuantLib_wrap.R instead of QuantLib.R. If using 1.3.31, you must uncomment the relevant sections in Makefile.am and makeRData.R.

Build Process: Run the make command to generate the following files:

• QuantLib.so (the R library)
• QuantLib_wrap.R (the wrapper script)
• QuantLib.RData (the compiled wrapper)

The build process requires that QuantLib headers and libraries are accessible to the compiler. Configuration varies by platform:

Unix-like platforms

• Ensure quantlib-config is in your PATH.
• Use CXXFLAGS and LDFLAGS to set appropriate compiler and linker flags.
• If Boost is required for compilation, set the INCLUDE environment variable to point to the Boost include directory.

Windows

• Define the QL_DIR environment variable pointing to your QuantLib directory (e.g., C:\Lib\QuantLib).
• If you are building against a debug version of QuantLib and intend to debug the library from Python, set the QL_DEBUG environment variable to TRUE.

To use the provided Visual Studio .NET projects, you must define an environment variable QL_DIR before launching the IDE. The value of QL_DIR must be the absolute path to your QuantLib installation directory.

Example value: C:\Lib\QuantLib

setx QL_DIR "C:\Lib\QuantLib"

Testing the built wheel requires tox. Install it via pip, then execute the test suite using tox run.

pip install tox
tox run

To run the examples on your local machine, you must install the required Python modules. The dependency list is maintained in the requirements.txt file located in the binder folder at the root of the repository. Navigate to that directory and use pip to install them.

pip install -r requirements.txt

If you encounter issues or have feedback regarding QuantLib-SWIG, you can use the following channels:

• GitHub Issues: Report bugs directly at the QuantLib-SWIG issues page.
• Mailing Lists: Use the quantlib-users and quantlib-dev mailing lists for general questions and feedback. Instructions for subscribing can be found at QuantLib Mailing Lists.

Once the library has been compiled using make, you can load it into your R session using the following sequence.

Important: You must call cacheMetaData(1) after loading to work around a bug that prevents S4 methods from loading correctly.

dyn.load('QuantLib.so')
load('QuantLib.RData')
cacheMetaData(1)

To build the Java bindings on a Linux system, you must provide the location of the JDK include directories during the configuration step. This allows the build system to locate the necessary Java headers.

Follow these steps:

1. Run ./autogen.sh to generate the configuration scripts.
2. Run ./configure with the --with-jdk-include and --with-jdk-system-include flags pointing to your JDK installation.
3. Run make to compile the module.

./autogen.sh
./configure --with-jdk-include=/usr/lib/jvm/java-1.5.0-sun-1.5.0.08/include \
  --with-jdk-system-include=usr/lib/jvm/java-1.5.0-sun-1.5.0.08/include/linux
make

You can execute the QuantLib Python examples in an interactive environment without local installation by using Binder. Click the 'Launch Binder' badge in the repository to open the examples in a hosted Jupyter environment.

https://mybinder.org/v2/gh/lballabio/QuantLib-SWIG/binder?urlpath=lab/tree/Python/examples

The .NET interfaces are compatible with Linux and macOS. To build them, navigate to the CSharp directory and execute the following sequence of commands after you have already completed the ./autogen.sh and ./configure steps in the main QuantLib-SWIG root directory:

1. Run make in the CSharp directory.