I have a unique system where customer IDs are linked with movie IDs in a dictionary. Even if the customer watches the same movie multiple times, I want to simplify it as a single entry. In order to achieve this, I need to convert my dictionary data into binary format. For each row, the customer IDs should be displayed and the columns should represent movie IDs. If the customer has watched the movie, it will show 1; otherwise, it will show 0.
d = {'121212121' : 111, 222, 333, 333,444, 444, '212121212' : 222, 555, 555, 666, '212123322' : 555, 666, 666, 666, 777}
Desired output :
customer ID 111 222 333 444 555 666 777
121212121 1 1 1 1 0 0 0
212121212 0 1 0 0 1 1 0
121323231 0 0 0 0 1 1 1
Previous attempts involved using count vectorizer()
Code :
cv = CountVectorizer()
movies = cv.fit_transform(cust['movies_list'])
cols = cv.vocabulary_
movies_ = pd.DataFrame(movies.toarray(), columns = cols, index =
cust['customer_id'])
movies_
Output :
customer ID 111 222 333 444 555 666 777
212121212 1 1 2 2 0 0 0
121212121 0 1 0 0 2 1 0
121323231 0 0 0 0 1 3 1
The issue was that the customer IDs did not align correctly, and instead of binary values, I received counts of how many times the movie was watched by each customer.