CODEWITHSUNDEEP
pandasconditional-statementscalculated-columns
frank
frank

Reputation: 3608

generate conditional column in pandas

I have:

pd.DataFrame({'price':['45p','£1.35']})

I want to convert these to numeric, and get:

pd.DataFrame({'price':['45p','£1.35'],'numeric':[0.45,1.35]})

I tried:

df['numeric']=np.where(df.price.str.contains('p') is True,
                  pd.to_numeric(df.price.str.replace('p',''))/100,
                 pd.to_numeric(df.price.str.replace('£','')))

and got the following error: ValueError: Unable to parse string "£1.35" at position 7

Any suggestions what I am doing wrong?

Upvotes: 1

Views: 16

Answers (1)

Quang Hoang
Quang Hoang

Reputation: 150745

Try one step at a time:

# where values in pennies
is_pennies = df['price'].str.contains('p')

# remove the currency characters and convert to numerics
df['price'] = df.price.str.replace('p|£', '').astype(float)

# update the values in pennies
df.loc[is_pennies, 'price'] /= 100

Output:

   price
0   0.45
1   1.35

Upvotes: 2

Related Questions