Version 5¶

Changes Since 5.5¶

• Removed dependence on property-cached

• Bumped minimum NumPy, SciPy, pandas, statsmodels and Cython

• Added compatability with Cython 3

Release 5.5¶

• NumPy 1.25 fixes

• Initial pandas copy-on-write support

• Switched doc theme to sphinx-immaterial

• Small fixes for typing issues

Release 5.4¶

• Compatability release with pandas 2.0

• Add testing and wheel support for Python 3.11

Release 5.3¶

• Fixed a bug in arch_model() where power was not passed to the FIGARCH constructor (GH572).

• Fixed a bug that affected downstream projects due to an overly specific assert (GH569).

Release 5.2¶

• Fixed a bug in in std_resid() that would raise an exception when the data used to construct the model with a NumPy array (GH565).

• Fixed a bug in forecast() and related forecast methods when producing multi-step forecasts usign simulation with exogenous variables (GH551).

Release 5.1¶

Unit Root¶

• Improved automatic lag length selection in DFGLS by using OLS rather than GLS detrended data when selecting the lag length. This problem was studied by Perron, P., & Qu, Z. (2007).

Release 5.0¶

Unit Root¶

• All unit root tests are now immutable, and so properties such as trend cannot be set after the test is created.

Bootstrap¶

• Added seed keyword argument to all bootstraps (e.g., IIDBootstrap and StationaryBootstrap) that allows a NumPy numpy.random.Generator to be used. The seed keyword argument also accepts legacy numpy.random.RandomState instances and integers. If an integer is passed, the random number generator is constructed by calling numpy.random.default_rng() The seed keyword argument replaces the random_state keyword argument.

• The random_state() property has also been deprecated in favor of generator().

• The get_state() and set_state() methods have been replaced by the state() property.

Volatility Modeling¶

• Added seed keyword argument to all distributions (e.g., Normal and StudentsT) that allows a NumPy numpy.random.Generator to be used. The seed keyword argument also accepts legacy numpy.random.RandomState instances and integers. If an integer is passed, the random number generator is constructed by calling numpy.random.default_rng() The seed keyword argument replaces the random_state keyword argument.

• The Normal.random_state property has also been deprecated in favor of generator().

• Added ARCHInMean mean process supporting (G)ARCH-in-mean models.

• Extended VolatilityProcess with volatility_updater() that contains a VolatilityUpdater to allow ARCHInMean to be created from different volatility processes.

Setup¶

• Added support for using an environmental variable to disable C-extension compilation.

  • Linux and OSX: export ARCH_NO_BINARY=1

  • PowerShell: $env:ARCH_NO_BINARY=1

  • cmd: set ARCH_NO_BINARY=1