Replacing NaN in floating point column with a blank, while maintaining data type

Question

I am trying to replace all NaN values in a dataframe with blanks. I have been able to replace the NaN values by using

df.fillna(\'\', inplace=True)