" A library made by quants for quants "
QuanticBoost is an integrated Python/C++/Cuda framework with the use of CMake and Docker containers. The library provides access to GPU computing, multi-threading tools, OpenMP and MPI libraries. The main goal is to offer a versatile environment to prototype and develop high-performance mathematical finance methods for derivatives pricing.
The scope of this library can cover Front Office, Market Risk, Counterparty Credit Risk and ALM Risk and potentially Machine Learning topics.
The library is structured as follows:
- Applications ( C++ and Python entry points )
- Libraries (C++, Cuda and python code)
- External Libraries (GoogleTests, Eigen, CppDB, Cub, CppAD, StanMath, AlgLib, QuantLib)
In addition to that, you are given a simple data model and local database with dummy inputs to fire up your calculations.
How to work with the library?
This repository offers a fully functionable virtual development infrastructure, for those you want to work with this library. In our wiki we explain how can you create virtual enviroments using Vagrant.
How to run on GPUs?
To run it, ideally you would need a GPU enabled workstation. In case that you do NOT have available GPUs, we launch an EC2 AWS instance using Vagrant AWS Plugin.
The design assumption
The library implements a quantitative analytics library called Calculation Engine on the diagram below.
The responsibilities of this engine is to cover the computational (mathematical) part of a wider Analytics
The library is able to interact with the rest of the application by receiving business requests to perform
quantitative calculations e.g.
- Present value, Sensitivities,
- General Exposure and XVA/PFE/Funding/Capital calculations
The code does not involve in any data plumbing and general IT activities, apart from using an external
Database API library (CppDb) to interact with the rest of the application. The code only assumes to receive
market and trade data in its own format, which is agnostic of any external business logic, data translations
or reconciliations. This part is supposed to be handled by the IT/DevOps nature of the final application.
How to use the documentation?
- The markdown files are used as high-level intro within the Bitbuckets pages. However, they can be viewed also locally from your workstation. In case you cannot view properly *.md files, there are approriate plugins that work smoothly on the popular IDEs or web browsers.
- The code is self-documented with Doxygen comments. Using the CMake option DOXYGEN_DOCUMENTATION you get the Doxygen pages as a part of the CMake build. So anything that has to do with the usage of classes and functions is there.
- Generic info, best practices and quidelines are available at the Bitbucket wiki.
Copyright © 2016 Panagiotis Nikolopoulos
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.