I need to do feature extraction using the column 'Amenities' from the dataframe house_price.
The column Amenities has the following set of data
house_data['Amenities']
3 3 beds 1 bath
4 1 bed 1 bath 1 parking
5 3 beds 1 bath
6 2 beds 2 baths 2 parking
7 3 beds 1 bath 2 parking
...
2096 3 beds 2 baths 1 parking 419m
2097 4 beds 1 bath 2 parking
2098 3 beds 2 baths 2 parking
2099 2 beds 2 baths 1 parking
2100 3 beds 2 baths 1 parking 590m
Name: Amenities, Length: 1213, dtype: object
I need to extract the number of beds, baths and parkings and store them into 3 seperate columns.
house_data["bedrooms"] = ''
house_data["bedrooms"] = house_data["Amenities"].str.extract("(\d*\.?\d+)", expand=True)
3 3
4 1
5 3
6 2
7 3
..
2096 3
2097 4
2098 3
2099 2
2100 3
Name: bedrooms, Length: 1213, dtype: object
The above code extracts only the first digit of the entire string. How can I extract the digits representing the number of baths/parking and store them under different columns?
We can use named groups here with Series.str.extract:
regex = r'(?P<beds>\d)\sbeds?\s(?P<bath>\d+)\sbaths?\s?(?P<parking>\d)?'
df = pd.concat([df, df['Amenities'].str.extract(regex)], axis=1)
Amenities beds bath parking
0 3 beds 1 bath 3 1 NaN
1 1 bed 1 bath 1 parking 1 1 1
2 3 beds 1 bath 3 1 NaN
3 2 beds 2 baths 2 parking 2 2 2
4 3 beds 1 bath 2 parking 3 1 2
5 3 beds 2 baths 1 parking 419m 3 2 1
6 4 beds 1 bath 2 parking 4 1 2
7 3 beds 2 baths 2 parking 3 2 2
8 2 beds 2 baths 1 parking 2 2 1
9 3 beds 2 baths 1 parking 590m 3 2 1