One hot encoder categorical columns NaN in Housing Dataset

Hello, when trying to apply the method to the categorical columns, I get an error message regarding the presence of NaN. I tries using the .replace(np.NaN, '', regex = True, inplace=True) method (and other similar methods) to convert them to blank cells, but the NaN stubbornly remains. Should the imputer be used on the Categorical data as well as the numerical data, as there is no obvious strategy to use?

Hey, as mentioned in the lecture, you can fill the values of NaN by making another category named "Unknown". You can use fillna() to do the same. This might create confusion so we will include this strategy in the assignment notebook soon.
Thanks! for reporting this.