v1.6.0

What's Changed

• Pandas objects (Series or DF) can now be passed as regression variables (y, x, yend, q, regimes), suppressing the need to explicitly provide their respective names in these cases. The names are automatically collected from the pandas objects. This applies to all regression functions.
• Implementation of Flexible WX. Through a list of Booleans in argument slx_vars, the user can now select which x variables are to be lagged. This new feature applies to OLS, TSLS, ML and GM spatial models (both lag and error).
• Kernel weights are now allowed in SLX and SDM models.
• Added Durbin-Wu-Hausman test on endogeneity of variables to TSLS.
• Spatial impacts are now computed for regression models with regimes.
• Added VIF method to OLS results.
• Added compatibility with libpysal.Graph.

Other Changes

• Fixing test errors by @pedrovma in #139
• Fixing more strange failing CI tests by @pedrovma in #140
• Updating example in test_sur.py by @pedrovma in #141
• update CI envs & testing workflow by @jGaboardi in #142
• modern micromamba - #143 by @jGaboardi in #144

Full Changelog: v1.5.0...v1.6.0