Skip to main content

The unofficial Python client for the Coinbase Pro API

Project description

coinbasepro-python

Build Status Coverage Status

The Python client for the Coinbase Pro API (formerly known as the GDAX)

Provided under MIT License by Daniel Paquin.

Note: this library may be subtly broken or buggy. The code is released under the MIT License – please take the following message to heart:

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Benefits

  • A simple to use python wrapper for both public and authenticated endpoints.
  • In about 10 minutes, you could be programmatically trading on one of the largest Bitcoin exchanges in the world!
  • Do not worry about handling the nuances of the API with easy-to-use methods for every API endpoint.
  • Gain an advantage in the market by getting under the hood of CB Pro to learn what and who is behind every tick.

Under Development

  • Test Scripts
  • Additional Functionality for the real-time order book
  • FIX API Client Looking for assistance

Getting Started

This README is documentation on the syntax of the python client presented in this repository. See function docstrings for full syntax details.
This API attempts to present a clean interface to CB Pro, but in order to use it to its full potential, you must familiarize yourself with the official CB Pro documentation.

pip install cbpro
#or
pip install git+git://github.com/yiwensong/coinbasepro-python.git

Public Client

Only some endpoints in the API are available to everyone. The public endpoints can be reached using PublicClient

import cbpro
public_client = cbpro.PublicClient()

PublicClient Methods

public_client.get_products()
# Get the order book at the default level.
public_client.get_product_order_book('BTC-USD')
# Get the order book at a specific level.
public_client.get_product_order_book('BTC-USD', level=1)
# Get the product ticker for a specific product.
public_client.get_product_ticker(product_id='ETH-USD')
# Get the product trades for a specific product.
# Returns a generator
public_client.get_product_trades(product_id='ETH-USD')
public_client.get_product_historic_rates('ETH-USD')
# To include other parameters, see function docstring:
public_client.get_product_historic_rates('ETH-USD', granularity=3000)
public_client.get_product_24hr_stats('ETH-USD')
public_client.get_currencies()
public_client.get_time()

Authenticated Client

Not all API endpoints are available to everyone. Those requiring user authentication can be reached using AuthenticatedClient. You must setup API access within your account settings. The AuthenticatedClient inherits all methods from the PublicClient class, so you will only need to initialize one if you are planning to integrate both into your script.

import cbpro
auth_client = cbpro.AuthenticatedClient(key, b64secret, passphrase)
# Use the sandbox API (requires a different set of API access credentials)
auth_client = cbpro.AuthenticatedClient(key, b64secret, passphrase,
                                  api_url="https://api-public.sandbox.pro.coinbase.com")

Pagination

Some calls are paginated, meaning multiple calls must be made to receive the full set of data. The CB Pro Python API provides an abstraction for paginated endpoints in the form of generators which provide a clean interface for iteration but may make multiple HTTP requests behind the scenes. The pagination options before, after, and limit may be supplied as keyword arguments if desired, but aren't necessary for typical use cases.

fills_gen = auth_client.get_fills()
# Get all fills (will possibly make multiple HTTP requests)
all_fills = list(fills_gen)

One use case for pagination parameters worth pointing out is retrieving only new data since the previous request. For the case of get_fills(), the trade_id is the parameter used for indexing. By passing before=some_trade_id, only fills more recent than that trade_id will be returned. Note that when using before, a maximum of 100 entries will be returned - this is a limitation of CB Pro.

from itertools import islice
# Get 5 most recent fills
recent_fills = islice(auth_client.get_fills(), 5)
# Only fetch new fills since last call by utilizing `before` parameter.
new_fills = auth_client.get_fills(before=recent_fills[0]['trade_id'])

AuthenticatedClient Methods

auth_client.get_accounts()
auth_client.get_account("7d0f7d8e-dd34-4d9c-a846-06f431c381ba")
# Returns generator:
auth_client.get_account_history("7d0f7d8e-dd34-4d9c-a846-06f431c381ba")
# Returns generator:
auth_client.get_account_holds("7d0f7d8e-dd34-4d9c-a846-06f431c381ba")
# Buy 0.01 BTC @ 100 USD
auth_client.buy(price='100.00', #USD
               size='0.01', #BTC
               order_type='limit',
               product_id='BTC-USD')
# Sell 0.01 BTC @ 200 USD
auth_client.sell(price='200.00', #USD
                size='0.01', #BTC
                order_type='limit',
                product_id='BTC-USD')
# Limit order-specific method
auth_client.place_limit_order(product_id='BTC-USD',
                              side='buy',
                              price='200.00',
                              size='0.01')
# Place a market order by specifying amount of USD to use.
# Alternatively, `size` could be used to specify quantity in BTC amount.
auth_client.place_market_order(product_id='BTC-USD',
                               side='buy',
                               funds='100.00')
# Stop order. `funds` can be used instead of `size` here.
auth_client.place_stop_order(product_id='BTC-USD',
                              side='buy',
                              price='200.00',
                              size='0.01')
auth_client.cancel_order("d50ec984-77a8-460a-b958-66f114b0de9b")
auth_client.cancel_all(product_id='BTC-USD')
# Returns generator:
auth_client.get_orders()
auth_client.get_order("d50ec984-77a8-460a-b958-66f114b0de9b")
# All return generators
auth_client.get_fills()
# Get fills for a specific order
auth_client.get_fills(order_id="d50ec984-77a8-460a-b958-66f114b0de9b")
# Get fills for a specific product
auth_client.get_fills(product_id="ETH-BTC")
depositParams = {
        'amount': '25.00', # Currency determined by account specified
        'coinbase_account_id': '60680c98bfe96c2601f27e9c'
}
auth_client.deposit(depositParams)
# Withdraw from CB Pro into Coinbase Wallet
withdrawParams = {
        'amount': '1.00', # Currency determined by account specified
        'coinbase_account_id': '536a541fa9393bb3c7000023'
}
auth_client.withdraw(withdrawParams)

WebsocketClient

If you would like to receive real-time market updates, you must subscribe to the websocket feed.

Subscribe to a single product

import cbpro
# Paramters are optional
wsClient = cbpro.WebsocketClient(url="wss://ws-feed.pro.coinbase.com", products="BTC-USD")
# Do other stuff...
wsClient.close()

Subscribe to multiple products

import cbpro
# Paramaters are optional
wsClient = cbpro.WebsocketClient(url="wss://ws-feed.pro.coinbase.com",
                                products=["BTC-USD", "ETH-USD"])
# Do other stuff...
wsClient.close()

WebsocketClient + Mongodb

The WebsocketClient now supports data gathering via MongoDB. Given a MongoDB collection, the WebsocketClient will stream results directly into the database collection.

# import PyMongo and connect to a local, running Mongo instance
from pymongo import MongoClient
import cbpro
mongo_client = MongoClient('mongodb://localhost:27017/')

# specify the database and collection
db = mongo_client.cryptocurrency_database
BTC_collection = db.BTC_collection

# instantiate a WebsocketClient instance, with a Mongo collection as a parameter
wsClient = cbpro.WebsocketClient(url="wss://ws-feed.pro.coinbase.com", products="BTC-USD",
    mongo_collection=BTC_collection, should_print=False)
wsClient.start()

WebsocketClient Methods

The WebsocketClient subscribes in a separate thread upon initialization. There are three methods which you could overwrite (before initialization) so it can react to the data streaming in. The current client is a template used for illustration purposes only.

  • onOpen - called once, immediately before the socket connection is made, this is where you want to add initial parameters.
  • onMessage - called once for every message that arrives and accepts one argument that contains the message of dict type.
  • on_close - called once after the websocket has been closed.
  • close - call this method to close the websocket connection (do not overwrite).
import cbpro, time
class myWebsocketClient(cbpro.WebsocketClient):
    def on_open(self):
        self.url = "wss://ws-feed.pro.coinbase.com/"
        self.products = ["LTC-USD"]
        self.message_count = 0
        print("Lets count the messages!")
    def on_message(self, msg):
        self.message_count += 1
        if 'price' in msg and 'type' in msg:
            print ("Message type:", msg["type"],
                   "\t@ {:.3f}".format(float(msg["price"])))
    def on_close(self):
        print("-- Goodbye! --")

wsClient = myWebsocketClient()
wsClient.start()
print(wsClient.url, wsClient.products)
while (wsClient.message_count < 500):
    print ("\nmessage_count =", "{} \n".format(wsClient.message_count))
    time.sleep(1)
wsClient.close()

Testing

A test suite is under development. Tests for the authenticated client require a set of sandbox API credentials. To provide them, rename api_config.json.example in the tests folder to api_config.json and edit the file accordingly. To run the tests, start in the project directory and run

python -m pytest

Real-time OrderBook

The OrderBook subscribes to a websocket and keeps a real-time record of the orderbook for the product_id input. Please provide your feedback for future improvements.

import cbpro, time
order_book = cbpro.OrderBook(product_id='BTC-USD')
order_book.start()
time.sleep(10)
order_book.close()

Testing

Unit tests are under development using the pytest framework. Contributions are welcome!

To run the full test suite, in the project directory run:

python -m pytest

Change Log

1.1.2 Current PyPI release

  • Refactor project for Coinbase Pro
  • Major overhaul on how pagination is handled

1.0

  • The first release that is not backwards compatible
  • Refactored to follow PEP 8 Standards
  • Improved Documentation

0.3

  • Added crypto and LTC deposit & withdraw (undocumented).
  • Added support for Margin trading (undocumented).
  • Enhanced functionality of the WebsocketClient.
  • Soft launch of the OrderBook (undocumented).
  • Minor bug squashing & syntax improvements.

0.2.2

  • Added additional API functionality such as cancelAll() and ETH withdrawal.

0.2.1

  • Allowed WebsocketClient to operate intuitively and restructured example workflow.

0.2.0

  • Renamed project to GDAX-Python
  • Merged Websocket updates to handle errors and reconnect.

0.1.2

  • Updated JSON handling for increased compatibility among some users.
  • Added support for payment methods, reports, and Coinbase user accounts.
  • Other compatibility updates.

0.1.1b2

  • Original PyPI Release.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

cbpro2-1.0.1.tar.gz (23.5 kB view details)

Uploaded Source

File details

Details for the file cbpro2-1.0.1.tar.gz.

File metadata

  • Download URL: cbpro2-1.0.1.tar.gz
  • Upload date:
  • Size: 23.5 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/1.12.1 pkginfo/1.4.2 requests/2.19.1 setuptools/40.4.3 requests-toolbelt/0.8.0 tqdm/4.27.0 CPython/2.7.14

File hashes

Hashes for cbpro2-1.0.1.tar.gz
Algorithm Hash digest
SHA256 4a34d01bc400b3a87892375aa9edb651faa8a935cfdc23dd8a52eb2626811e2c
MD5 65df893ea9684819ea99069c7c45c779
BLAKE2b-256 168f6f7b17fc7b50ba31e8cef42959138b7aad23f84643ede46da16814788261

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page