Awesome Open Source
Awesome Open Source



KISS - Keep it simple, stupid!

Algorithm script file ( ...

import logging
log = logging.getLogger("friar_tuck")
def initialize(context, data):
    # (required) is called when the process starts up
    # Note: within this method "initialize", the parameter "data" should primarily be used to load historical data for initialization, the use of data.current(...) method is best within the handle_data(...) method."hello, I am in initialize...")
def on_market_open(context, data):
    # (optional) is called when the market opens or after a restart of the process during the live market
    # Note: within this method "on_market_open", the parameter "data" should primarily be used to load historical data for initialization, the use of data.current(...) method is best within the handle_data(...) method."on market open")
def handle_data(context, data):
    # (required) is called at each data interval, currently supported frequencies (1m=every minute, 1h=every hour, 1d=every day at end of session)."hello, I am in handle_data")

Config file (rh_config.cfg) ...

    username = <robinhood username>
    password = <robinhood password>
    apikey = <get_your_fee_apikey from>
    wait_for_connection = yes

Run FriarTuck - Live

python --algo_script algo_template --config_file rh_config.cfg --data_frequency 1h


Drive for this project: Build an algorithm on Quantopian, was satisfied with the results, the day I was ready to run live I found out that Quantopian was discontinuing live trading... Bummer!!! Quantopian was my only hope for trading algorithmic on Robinhood and thought all hopes were gone. Than I came across the unofficial api for Robinhood (, regained some hopes again... Spent sometime leveraging the Jamonek's Python implementation of the API to build a framework similar to Quantopian. Since Quantopian is keeping their framework for backtesting; my aim for this framework is to keep it was close to Quantopian as possible, this way I can backtest on Quantopian and run live using this framework.

Using the following projects/data (Respect!):
    # As a broker to Robinhood
    # IEXTRADING for historical data
    # ALPHA VANTAGE for intra-day realtime data (5 requests per minute for free, get your free apikey... more request are offered for premium service)



  • ROBINHOOD Trading Account
  • An environment to run FriarTuck (All my testing so far has been on Windows)
  • Python (built this using version 3.4)


pip install -r requirements.txt


Load Contracts

    context.aapl = lookup_security("AAPL")
    context.wtw = lookup_security("WTW") = lookup_security("FIT")
    # Get active trading account details
    my_account = context.account
    # Get active portfolio details
    my_portfolio = context.portfolio
    # Iterate through all open positions
    for security in context.portfolio.positions:"symbol(%s) pos(%s) " % (security.symbol, context.portfolio.positions[security]))

Using "data" object for current data from "handle_data(context, data)"

    # Get current data (all fields [open, high, low, close, volume, price, bid_price, bid_size, ask_price, ask_size])
    current_quote = data.current(context.aapl)
    # Get current data (specific fields)
    current_quote = data.current(context.aapl, field=['close', 'open'])
    # Get field(s) for more than 1 security
    current_quote = data.current([context.aapl, context.wtw], field='close')

Using "data" object for historical-data from "handle_data(context, data) and on_market_open(context, data)"

    hist_quotes = data.history([context.aapl, context.wtw], frequency='1m', bar_count=10, field='close')

    current_data = data.current(, field=['close', 'price'])

Sample code using pandas to load contracts from external file

    import pandas as pd
    context.assets = []
    context.symbol_metadata = {}
    dataset = pd.read_csv("")
    for (index, series) in dataset.iterrows():
        asset = lookup_security(series["symbol"])
        context.symbol_metadata[asset] = series
    log.debug("symbol_metadata (%s)" % context.symbol_metadata)

Ordering using Friar Tuck

    # Ordering using monetary value; this will use last trade price to calculate number of shares to order
    # for buy
    cash = 1000
    # or for sell
    cash = -1000 
    context.aapl_order_id = order_value(context.aapl, cash, order_type=OrderType(stop_price=158.60), time_in_force='gtc')
    # Ordering a set number of shares
    # for buy
    shares = 120
    # or for sell
    shares = -120
    context.aapl_order_id = order_shares(context.aapl, shares, order_type=OrderType(stop_price=158.60), time_in_force='gtc')
    # retrieve order object
    order = get_order(context.aapl_order_id)"order=%s" % order)
    last_filled_buy_order = get_last_filled_buy_order(context.aapl)
    last_filled_sell_order = get_last_filled_sell_order(context.aapl)
    last_filled_orders_by_side = get_last_filled_orders_by_side(context.aapl)
    last_filled_buy_order = last_filled_orders_by_side["buy"]
    last_filled_sell_order = last_filled_orders_by_side["sell"]
    # Retrieve all open orders
    open_orders = get_open_orders()
    for stock_symbol in open_orders:"open_order=%s" % open_orders[stock_symbol])
    # Retrieve all open orders by security
    open_orders = get_open_orders(context.aapl.symbol)
    for open_order in open_orders:"AAPL open_order=%s" % open_order)
    # Cancel an order


WARNING!: This api is subject to change! Make sure you know what you're doing!

Much of this API is the same of similar to Quantopian

Context Object:
    is_market_open (boolean): Indicates if the stock market is open or not.
    account (Account Object): Contains the details about the trading account.
    portfolio (Portfolio Object): holds portfolio details including positions.
Security Object: Used when interacting with the broker.
    symbol (string): Stock symbol
    simple_name (string): User-friendly security name
    min_tick_size (float): minimum tick-size of the production, important when supplying a price for an order.
    is_tradeable (boolean): Indicates if the security is tradeable
    security_type (String): The security type
    security_detail (json object): The actual security details from Robinhood
OrderType Object: Use to determine the type of order (limit, stop, stop-limit)
    price (float): Use to determine a limit order
    stop_price (float): Use to determine a stop order
    Note: both fields can be set from the constructor (price=my_limit_price, stop_price=my_stop_price)... 
    If both fields are obmitted then its a market order, however, 
    please note that though it's a market order to you but robinhood always uses a limit order
Order Object: This represents an order on the market
    id (string): this is the unique identifier of the order
    status (integer): The status of the order
        0 = Open (Robinhood confirmed/partially_filled)
        1 = Filled
        2 = Cancelled
        3 = Rejected
        4 = Unconfirmed (Robinhood queued/unconfirmed)
        5 = Failed
    created (datetime): the date and time the order was created
    stop (float): Stop price
    limit (float): Limit price
    amount (int): Shares ordered
    symbol (string): Stock symbol
    filled (boolean): Indicates if the entire order is filled
    commission (float): commissione charged
    rejected_reason (string): if the order was rejected, this will be the reason
    time_in_force (string): (Robinhood [gfd|gtc])
Portfolio Object:
    capital_used (float): net capital at play (Total cost of shorts) minus (Total cost of longs)
    cash (float): total cash at hand available for trading (robinhood unsettled cash + cash)
    pnl (float): net profit_loss (equity - uncleared_deposits - equity_previous_close)
    positions (dict): key=Security object, value=Position object
    portfolio_value (float): total value of the portfolio (robinhood unsettled cash + cash + market-value)
    positions_value (float): market_value (Robinhood)
    returns (float): total returns since starting the FriarTuck process ((portfolio_value - starting_cash) / starting_cash)
    starting_cash (float): total available cash at the start of the FriarTuck process
    start_date (datetime): The date and time the FriarTuck process started
Position Object: (context->portfolio->positions)
    amount (int): The number of shares
    cost_basis (float): The average price of the position.
    last_sale_price (float): The more recent quote trade price for the security.
    created (datetime): create datetime of the position
Account Object: These fields are from Quantopian, I attempt to map them with fields from and calcs from Robinhood
    accrued_interest (float): This will always be with Robinhood(could not find a matching field) 0
    available_funds (float): cash available for trading (Robinhood unsettled cash + cash)
    buying_power (float): value available to buy stock (equity - market_value - cash_held_for_orders)
    cushion (float): (Robinhood unsettled cash + cash) / portfolio_value
    day_trades_remaining (int): Infinity (could not identify a Robinhood field to match)
    equity_with_loan (float): portfolio_value
    excess_liquidity (float): (Robinhood unsettled cash + cash)
    initial_margin_requirement (float): 0 (could not identify a Robinhood field to match)
    leverage (float): current leverage = (long_position_value + short_position_value) / portfolio_value
    maintenance_margin_requirement (float): 0 (could not identify a Robinhood field to match)
    net_leverage (float): market_value / portfolio_value
    net_liquidation (float): portfolio_value
    regt_equity (float): (Robinhood unsettled cash + cash)
    regt_margin (float): Infinity (could not identify a Robinhood field to match)
    settled_cash (float): cash (robinhood cash)
    total_positions_value (float): (Robinhood market_value)


You can follow me on twitter.


I want to praise the efforts of the people/projects that have inspired me while
I've been working on this project by briefly mention below their names and projects:


Check out license for more details.


This project is currently in the very alpha phase, its changing all the time; production readiness is fully determined by you. I can not guarantee that this project has no bugs, I do try to minimize them but don't have the time or the resources eliminate them. I am testing this project live as I am using it for my own use, however, I am afraid that my testing is and will be bias since I'll somehow shape it to solve only my problems. To broaden this up, it would be nice to expand this project to solve other quant issues as well. Looking for volunteers to test and to develop this project for others to use. Use of this product is entirely AT YOUR OWN RISK; I do not and can not guarantee any part of this code.

If you want to help me please follow here.

Go back to the project description

Copyright © 2017 Code Society

Get A Weekly Email With Trending Projects For These Topics
No Spam. Unsubscribe easily at any time.
Python (1,136,168
Time Series (1,919
Trading (1,456
Stock (1,285
Broker (265
Quant (230
Stock Prices (185
Related Projects