Replace dots in a float column with nan in Python

Question

I have a data frame df like this

df = pd.DataFrame([
    {'Name': 'Chris', 'Item Purchased': 'Sponge', 'Cost': 22.50},
    {'Name': 'Kevyn', 'Item Purchased': 'Kitty Litter', 'Cost': '.........'},
    {'Name': 'Filip', 'Item Purchased': 'Spoon', 'Cost': '...'}],
    index=['Store 1', 'Store 1', 'Store 2'])

I want to replace the missing values in 'Cost' columns to np.nan. So far I have tried:

df['Cost']=df['Cost'].str.replace("\.\.+", np.nan)

and

df['Cost']=re.sub('\.\.+',np.nan,df['Cost'])

but neither of them seem to work properly. Please help.

Answer 1

Use DataFrame.replace with the regex=True switch.

df = df.replace('\.+', np.nan, regex=True)
df

         Cost Item Purchased   Name
Store 1  22.5         Sponge  Chris
Store 1   NaN   Kitty Litter  Kevyn
Store 2   NaN          Spoon  Filip

The pattern \.+ specifies one or more dots. You could also use [.]+ as a pattern to the same effect.