az.summary is one of the most commonly used functions in ArviZ. It displays a dataframe with summary statistics and/or diagnostics. Given that summary statistics are computed differently for circular variables, this function had a boolean argument for handling them: include_circ. When True, az.summary computed circular statistics for every variable in the model and displayed them on extra columns along the non-circular statistics dataframe. In the past few days, I changed this argument to circ_var_names, and now takes an iterable object. With this argument, the user can indicate which of the variables in the model are circular. Consequently, az.summary computes circular statistics only for said variables with no need to append extra columns. You can take a look at the code in this PR. See you in the next post! :)