按 Python 中的特定字符将一列分成两列

Question

I use Python3 and need to split price column which mixed price_value and price_unit together in a dataframe, the example data looks like 20dollar/m2/month or 1.8dollar/m2/day , I want split them to this format by word dollar :

price_value      price_unit
20             dollar/m2/month
1.8            dollar/m2/day

I have tried with the following code:

Option 1:

df['price_value'] = df['price'].apply(lambda row: row.split('dollar')[0])
df['price_unit'] = df['price'].apply(lambda row: row.split('dollar')[-1])

Option 2:

df['price_value'], df['price_unit'] = df1["price"].str.split('dollar', 1).str

But I get:

price_value      price_unit
20                /m2/month
1.8               /m2/day

How can I split them correctly? Thanks.

Answer 1

You may use str.extract with a r'(?P<price_value>.*?)(?P<price_unit>dollar.*)' regex:

>>> import pandas as pd
>>> df = pd.DataFrame(data=['20dollar/m2/month', '1.8dollar/m2/day'], columns=['price'])
>>> df['price'].str.extract(r'(?P<price_value>.*?)(?P<price_unit>dollar.*)')
  price_value       price_unit
0          20  dollar/m2/month
1         1.8    dollar/m2/day

See the regex demo .

Details

• (?P<price_value>.*?) - Group "price_value": any 0+ chars other than line break chars as few as possible
• (?P<price_unit>dollar.*) - Group "price_unit": dollar and any 0+ chars other than line break chars as many as possible.

I assume that you do not have any line breaks in the input, but if you happen to have any, prepend the pattern with the inline DOTALL modifier, (?s) : r'(?s)(?P<price_value>.*?)(?P<price_unit>dollar.*)'

To add the newly extracted columns to the existing data frame, you may also use

df[['price_value', 'price_unit']] = df['price'].str.extract(r'(.*?)(dollar.*)')

Here, named capturing groups are not necessary since you define the column names beforehand.

Answer 2

You could do:

df = pd.DataFrame(data=['20dollar/m2/month', '1.8dollar/m2/day'], columns=['price_unit'])

# split by capture group
result = df['price_unit'].str.split('(dollar.*$)', expand=True).drop(2, axis=1)

# rename columns
result.columns = ['price_value', 'price_unit']

print(result)

Output

  price_value       price_unit
0          20  dollar/m2/month
1         1.8    dollar/m2/day