Understanding how to query the book efficiently about hft-orderbook HOT 3 CLOSED

Hey @kjgd ,
sorry for the delayed answer, the holidays are a busy time at my house :)

Just to clarify - you want to query the top X price levels on the bid and ask sides, right?

try this:

levels_sorted = sorted(lob._price_levels.keys())
bids = [price_level for price_level in levels_sorted if price_level < lob.best_ask]
asks = [price_level for price_level in levels_sorted if price_level > lob.best_bid]

That gives you the order book's limit levels as sorted dict keys.

Unfortunately, we are dealing with a binary tree, so trying to find the "next lowest" or "next highest" price might mean going all the way down to a leaf node... hardly efficient either...

this Implementation is meant to keep track of an order book in an efficient way - not display it efficiently. You're therefore correct that it doesn't do that very well.

from hft-orderbook.

Hey @nlsdfnbch no worries :)

Yes, exactly. Your way is considerably neater, thanks.

this Implementation is meant to keep track of an order book in an efficient way - not display it efficiently. You're therefore correct that it doesn't do that very well.

I see. It seems an unusual design choice... how is the implementation meant to be used, then?

from hft-orderbook.

It was initially meant as the core component of a trading engine. The main goal was to be able to quickly determine top levels and execute trades. We never really thought about visualizing the book.

from hft-orderbook.