0.5.1

pushed to PYPI with correct C++ header

0.5

uses quadprog>=0.1.12.

v0.4

First official release to support Windows.

v0.3

• added faster method by Choi 2022 which improves upon the CCD algorithm by Spinu.
• the vanilla function now accepts an argument called method, whose available values are "spinu" (default, for backwards compatibility reasons) and "choi" for Choi's method.

v0.2

• using jax instead of tensorflow for automatic differentiation

v0.1.6

• removed property variance in favour of volatility in portfolio objects.

v0.1.1

Some macOS users reported having difficulties installing the library. More precisely, they were facing errors like
gcc: error: unrecognized command line option ‘-stdlib=libc++’.

This release hopefully fixes these installation issues.

Thanks,
Zé Vinícius.

v0.0.8

In this release with added support to:

• general linear equality and inequality constraints
• optimize for the mean return
• optimize for the volatility

The support for general linear constraints is done by passing the constraints matrices directly to the off-the-shelf QP solver. In the next release we will add scalable algorithms for this scenario, as it is done in https://github.com/dppalomar/riskParityPortfolio.

Tutorials are available at https://mirca.github.io/riskparity.py.

As always, if you have any questions, feel free to open an issue at https://github.com/mirca/riskparity.py/issues or send us an email at [email protected].

Thanks,
Zé