Financial Modelling in Python

P1: JYS

c03 JWBK378-Fletcher May 1, 2009 19:52 Printer: Yet to come

3

Extending Python from C++

It is usual in financial institutions that make use of quantitative analysis programs to have a

considerable investment in C++. Thus it can be important to foster interoperability between

C++ and Python. This chapter studies how Python modules can be implemented in C++ by

means of the Boost.Python

1

library (see also Appendix B for a primer on the Boost.Python

library).

3.1 BOOST.DATE TIME TYPES

It is common in quantitative analysis programming to require manipulation of and computa-

tions involving dates. The ‘Python Library’ contains excellent functionality for such activities.

Pricing systems written in C++, however, will be implemented using C++ datatypes for the

representation of dates and times. For pricing frameworks implemented in a hybrid of Python

and C++, it would be convenient to settle on a common representation of these fundamental

types. Accordingly, in this section we demonstrate the ‘reflection’ of functionality from the

C++ Boost.Date

Time library to Python.

Our reflection of the C++ date types into Python will be housed in the Python module

‘ppf

date time.pyd’, implemented in C++. We declare this intention in the entry point to our

Python module in the file ‘module.cpp’:

#include <boost/python/module.hpp>

namespace ppf

{

namespace date

time

{

void register

date();

void register

date more();

} // namespace ...