a2rl.DiscreteTokenizer.check_numerical_columns#

DiscreteTokenizer.check_numerical_columns(df)[source]#

Input validation on the all-numerical input dataframe.

Each column in the input dataframe must contain only finite values, and it cannot have just a single unique value. Callers are responsible to ensure the input dataframe contains only numeric columns.

These are considered non-finite values: None, numpy.nan, numpy.inf, pandas.NA (i.e., the nullable integers).

Parameters:: df (DataFrame) – an input dataframe whose all columns must be numeric. Callers must ensure to pass an all-numerical input dataframe.
Return type:: None