similar_records.py

from difflib import SequenceMatcher

def get_similar_records(df, column, input_string, num_records):
    # Create an empty list to store the similar records
    similar_records = []
    
    # Loop through the records in the column
    for record in df[column]:
        # Calculate the similarity between the input string and the record
        similarity = SequenceMatcher(None, input_string, record).ratio()
        
        # Add the record and its similarity to the list
        similar_records.append((record, similarity))
    
    # Sort the list of records by similarity in descending order
    similar_records.sort(key=lambda x: x[1], reverse=True)
    
    # Return the top num_records records
    return similar_records[:num_records]