df.isna().sum() not counting nan values

Question

I have several values in a Dataframe like this:

The nan values in "Flat_levels" are coming from:

df[\'Flat_level\'] = df[\'Flat_level