python convert nan to empty string

# Option 1
df1 = df.replace(np.nan, '', regex=True) # All data frame

# Option 2
df[['column1','column2']] = df[['column1','column2']].fillna('') # Specific columns

df.fillna('', inplace=True)