Creating a new column by mapping a dictionary to a data frame with partial string matching

Question

I'm trying to create a new column in a pandas dataframe based if a column contains a partial string.

I have a simple dataframe

    Food    Price
0   apple   1.00
1   banana  2.99
2   carrot  3.50

and I would like to append a column to it like so

        Food     Price  Sale Status
0   apple    1.00   on sale
1   banana   2.99   not on sale
2   carrot   3.50   on sale next week

This work if my keys are exact matches to the values in the Food column

my_dict={'apple':'on sale', 'banana':'not on sale', 'carrot':'on sale next week'}

df['Sale Status']= df['Food'].map(my_dict)

My problem is that my dataframe looks like this

    Food                Price
0   some other words apple  1.00
1   other banana text   2.99
2   blah blah carrot    3.50

Is there a way to match the dictionary values if the sting in Food contains the key name?

Answer 1

There are several solutions. You can iterate over 'Food' and map the key to its value if found, like

df['Sale Status'] = df['Food'].apply(lambda x: [v for k, v in my_dict.items() if k in x][0])

Or you can use str.extract()

df['Sale Status'] = df['Food'].str.extract('('+'|'.join(my_dict.keys())+')', expand=False).map(my_dict)