Orderbook

A fast L2/L3 orderbook data structure, in C, for Python

Basic Usage

from decimal import Decimal

import requests
from order_book import OrderBook

ob = OrderBook()

# get some orderbook data
data = requests.get("https://api.pro.coinbase.com/products/BTC-USD/book?level=2").json()

ob.bids = {Decimal(price): size for price, size, _ in data['bids']}
ob.asks = {Decimal(price): size for price, size, _ in data['asks']}

# OR

for side in data:
    # there is additional data we need to ignore
    if side in {'bids', 'asks'}:
        ob[side] = {Decimal(price): size for price, size, _ in data[side]}


# Data is accessible by .index(), which returns a tuple of (price, size) at that level in the book
price, size = ob.bids.index(0)
print(f"Best bid price: {price} size: {size}")

price, size = ob.asks.index(0)
print(f"Best ask price: {price} size: {size}")

print(f"The spread is {ob.asks.index(0)[0] - ob.bids.index(0)[0]}\n\n")

# Data is accessible via iteration
# Note: bids/asks are iterators

print("Bids")
for price in ob.bids:
    print(f"Price: {price} Size: {ob.bids[price]}")


print("\n\nAsks")
for price in ob.asks:
    print(f"Price: {price} Size: {ob.asks[price]}")


# Data can be exported to a sorted dictionary
# In Python3.7+ dictionaries remain in insertion ordering. The
# dict returned by .to_dict() has had its keys inserted in sorted order
print("\n\nRaw asks dictionary")
print(ob.asks.to_dict())

Main Features

• Sides maintained in correct order
• Can perform orderbook checksums
• Supports max depth and depth truncation

Installation

The preferable way to install is via pip - pip install order-book. Installing from source will require a compiler and can be done with setuptools: python setup.py install.

Running code coverage

The script coverage.sh will compile the source using the -coverage CFLAG, run the unit tests, and build a coverage report in HTML. The script uses tools that may need to be installed (coverage, lcov, genhtml).

Running the performance tests

You can run the performance tests like so: python perf/performance_test.py. The program will profile the time to run for random data samples of various sizes as well as the construction of a sorted orderbook using live L2 orderbook data from Coinbase.

The performance of constructing a sorted orderbook (using live data from Coinbase) using this C library, versus a pure Python sorted dictionary library:

The performance of constructing sorted dictionaries using the same libraries, as well as the cost of building unsorted, python dictionaies for dictionaries of random floating point data:

This represents a roughly 2x speedup compared to a pure python implementation, and in many cases is close to the performance of an unsorted python dictionary.

For other performance metrics, run performance_test.py as well as the other performance tests in perf/