Simulating scikit-learn’s OneHotEncoder mixed data type error My focus is on the caveat mentioned in OneHotEncoder’s list parameter doc.
