data['horsepower'].replace(to_replace='?' , value = np.nan,inplace = True) data['horsepower'].unique()
df['salary'] = df['salary'].fillna(df['salary'].mode()[0])