Suppose you have a stream of stock events with a requirement to retrieve the top K aggregated summations by stock symbol in descending order for only the last 20 seconds. Each event is a tuple with a time, stock symbol, and value. Timestamps are guaranteed to be in increasing order, but not unique. You have can have multiple events come with the same timestamp. What\'s the most optimal way of doing this? The brute force approach would be to recompute the aggregates each time the retrieve function is called, so you can account for events that dropped off the 20 second window.\n\nFor example\nApple 1s 10\nApple 2s 10\nTesla 5s 10\nGoogle 10s 50\nApple 19s 15\n--------- retrieve top 2 called ---- (Google, 50), (Apple, 35)\nTesla 25s 20\n----- retrieve top 2 called -------(Google, 50), (Tesla, 30)

Suppose you have a stream of stock events with a requirement to retrieve the top K aggregated summations by stock symbol in descending order for only the last 2