# Welcome to the Kernel Quantum Probability (KQP) library

This project aims at providing an API that allows to compute quantum densities (semi-definite positive hermitian operators) or events (subspaces).

It provides tools to compute quantum probabilities and update densities (conditionalisation), and supports the use of kernels to implicitly define the space, thus allowing working in very high dimensional spaces.

If you use KQP, please cite the **associated technical report**

Features

- L-GPL license
- Wrappers in Java and Python
- Various methods to compute a symmetric kernel EVD, i.e. $$\sum_i \alpha_i \varphi_i \varphi_i^\dagger $$ where \(\varphi_i\) lies in an (implicit) feature space
- Various feature spaces:
- Finite Hilbert space: Dense, Sparse, Sparse with dense subspace
- Kernels: Gaussian, polynomial, linear combination

- Quantum probabilities:
- computation,
- conditionalisation,
- entropy,
- and divergence.
- partials with respect to kernel parameters

- Other modules:
- Cone Quadratic Optimization (translated from python from cvxopt),
- Pre-image removal,
- rank-one fast EVD update

## Installation and usage

To install KQP, please refer to this page

### C++

### Java

### Python

## News

24 Apr 2013 – Learning kernel parameters

24 Apr 2013 – JSON data

07 May 2012 – First public release