quantpylib.hft.mocks

The mocks module contains classes that replicate the quantpylib.hft.feed.Feed and quantpylib.hft.oms.OMS trading agents, as well as the Replayer class that drives these mock classes. The Replayer has auxiliary statistical and graphical utility functions.

Latencies

Latency types/constants.

MockFeed

Bases: Feed

add_l2_book_feed(exc, ticker, handler=None, depth=20, buffer=100, **kwargs) async

Add a level 2 book feed.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required
ticker	str	The ticker.	required
handler	coroutine	The handler to add. Defaults to None.	None
depth	int	The depth of the order book. Defaults to 20.	20
buffer	int	The buffer size. Defaults to 100.	100

add_trades_feed(exc, ticker, handler=None, buffer=100, **kwargs) async

Add a trades feed.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required
ticker	str	The ticker.	required
handler	coroutine	The handler to add. Defaults to None.	None
buffer	int	The buffer size. Defaults to 100.	100

MockOMS

Bases: OMS

add_clock_callback(callback, name='CLOCK', interval_ms=500) async

Add a callback to the event queue that is triggered every interval_ms. Can be used for clock-based trading agents.

Parameters:

Name	Type	Description	Default
callback	coroutine	The callback to trigger.	required
name	str	The name of the callback. Defaults to 'CLOCK'.	'CLOCK'
interval_ms	int	The interval in milliseconds. Defaults to 500.	500

get_positions_value(exc)

Get the net signed value of all open positions for an exchange at replay time.

lagged_price(exc, ticker)

Get the price for a given exchange and ticker at current replay time.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required
ticker	str	The ticker.	required

ReplayEvents

Replay event types/constants.

Replayer

__init__(l2_data, trade_data, maker_fees=0.0001, taker_fees=0.0003, gateway=None, exchanges=None, latencies={Latencies.REQ_PUBLIC: 100, Latencies.REQ_PRIVATE: 50, Latencies.ACK_PUBLIC: 100, Latencies.ACK_PRIVATE: 50, Latencies.FEED_PUBLIC: 100, Latencies.FEED_PRIVATE: 50}, exchange_fees=None, exchange_latencies=None, stamp_cycle=1000)

A replayer for backtesting. This class replays orderboook depth data, trade data and provides a mock feed and oms classes to replay agent-based communications between the trading logic and the exchange. Simulates data feed, order submissions, matching, latencies and more.

Parameters:

Name	Type	Description	Default
l2_data	dict	Level 2 data mapping {exc:{ticker:[{ts,b,a}]}}.	required
trade_data	dict	Trade data mapping {exc:{ticker:[(ts,price,sz,dir)]}}.	required
maker_fees	float	The maker fees. Defaults to 0.0001.	0.0001
taker_fees	float	The taker fees. Defaults to 0.0003.	0.0003
gateway	`quantpylib.gateway.master.Gateway`	The gateway used to instantiate contract rules. Defaults to None.	None
exchanges	list	The list of exchanges. Defaults to keys inside the l2_data dict.	None
latencies	dict	Latency mapping in ms for each message channel. Defaults to 50 for all private and 100 for public communications. Keys are Latencies.REQ_PUBLIC, Latencies.REQ_PRIVATE, Latencies.ACK_PUBLIC, Latencies.ACK_PRIVATE, Latencies.FEED_PUBLIC, Latencies.FEED_PRIVATE.	{REQ_PUBLIC: 100, REQ_PRIVATE: 50, ACK_PUBLIC: 100, ACK_PRIVATE: 50, FEED_PUBLIC: 100, FEED_PRIVATE: 50}
exchange_fees	dict	The exchange fees specify fees dict by exchange. Defaults to maker_fees and taker_fees for each exchange.	None
exchange_latencies	dict	The exchange latencies specify latencies dict by exchange. Defaults to latencies for each exchange.	None
stamp_cycle	int	The interval in ms used for stamping backtest statistics such as portfolio equity, inventory. Defaults to 1000.	1000

add_callback(event_id, callback)

Add a callback to an event.

Parameters:

Name	Type	Description	Default
event_id	str	The event id.	required
callback	coroutine	The callback for events belonging to event_id.	required

add_events(event_id, timestamps, events, exc, latency=Latencies.NULL)

Add events to the replayer queue.

Parameters:

Name	Type	Description	Default
event_id	str	The event id.	required
timestamps	list	The timestamps of event occurence.	required
events	list	The events. Note len(timestamps) == len(events).	required
exc	str	The exchange.	required
latency	str	The latency for event arrival added to the timestamps of event occurence. Defaults to Latencies.NULL, zero latency.	NULL

add_to_queue(event_id, timestamp, msg, exc, latency=Latencies.NULL)

Add an event to the replayer queue.

Parameters:

Name	Type	Description	Default
event_id	str	The event id.	required
timestamp	int	The timestamp of event occurence.	required
msg	Any	The message/event.	required
exc	str	The exchange.	required
latency	str	The latency for event arrival added to the timestamps of event occurence. Defaults to Latencies.NULL, zero latency.	NULL

augment_forward_prices(df, forward_intervals=[0, 1, 10, 30, 60, 120, 300])

Augment the dataframe with forward prices for each interval. df should contain columns ts,exc,ticker. to match forward prices.

declare(key, value, exc=None, ticker=None, group=None)

Declare a value for logging purposes. Can later be used for tick-data modelling and analysis.

Parameters:

Name	Type	Description	Default
key	str	The key.	required
value	Any	The value.	required
exc	str	The exchange. Defaults to None.	None
ticker	str	The ticker. Defaults to None.	None
group	str	The group this key belongs to. Defaults to None.	None

df_actions(exc=None, ticker=None)

Get a dataframe of all order fills, with columns ts,price,size,dir,is_maker,exc,ticker.

df_counterparty_pnl(exc_tickers, plot=True, show=True, save=None)

Get the normalized counterparty pnl comparison for mapping between exchange and list of tickers. This simulates being the maker to all market orders that are traded on the exchange's market with zero fees.

df_declarations(key=None, exc=None, ticker=None, group=None)

Retrieve the declarations logs as a dataframe.

df_exchange(exc, plot=True, show=True, save=None)

Get the equity and inventory breakdown for an exchange.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required
plot	bool	Plot the pnl and inventory breakdown. Defaults to True.	True
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None

df_fairprices(group, key=None, exc=None, ticker=None, forward_intervals=[0, 1, 10, 30, 60, 120, 300], plot=False, show=True, save=None)

If a group declared in Replayer.declare is a fairprice estimator, this method performs a fairprice analysis and metrics comparison by common objectives such as MSE, MAD, etc by comparing the declared fairprice with the actual price at different forward intervals.

Parameters:

Name	Type	Description	Default
group	str	The group.	required
key	str	The key. Defaults to None.	None
exc	str	The exchange. Defaults to None.	None
ticker	str	The ticker. Defaults to None.	None
forward_intervals	list	The forward intervals in seconds. Defaults to [0,1,10,30,60,120,300].	[0, 1, 10, 30, 60, 120, 300]
plot	bool	Plot the results. Defaults to False.	False
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None

df_fills(ticker, exc, plot=True, ax=None, show=True, save=None)

Get fills for a ticker and exchange.

Parameters:

Name	Type	Description	Default
ticker	str	The ticker.	required
exc	str	The exchange.	required
plot	bool	Plot the fills. Defaults to True.	True
ax	Axes	The axes to plot on. Defaults to None.	None
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None

df_markouts(ticker=None, exc=None, plot=True, show=True, save=None, intervals=list(np.linspace(1, 151, 31) * 1000))

Get markouts for a ticker and exchange.

Parameters:

Name	Type	Description	Default
ticker	str	The ticker. Defaults to None. If None, all tickers are aggregated.	None
exc	str	The exchange. Defaults to None. If None, all exchanges are aggregated.	None
plot	bool	Plot the markouts. Defaults to True.	True
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None
intervals	list	The intervals in ms. Defaults to list(np.linspace(1, 151, 31) * 1000).	list(linspace(1, 151, 31) * 1000)

df_portfolio(plot=True, show=True, save=None)

Get the portfolio pnl breakdown by exchange.

Parameters:

Name	Type	Description	Default
plot	bool	Plot the portfolio pnl breakdown. Defaults to True.	True
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None

df_prices(ticker, exc, plot=True, ax=None, with_fill=True, with_inventory=None, key=None, group=None, share_index=False, show=True, save=None)

Get tick prices for a ticker and exchange.

Parameters:

Name	Type	Description	Default
ticker	str	The ticker.	required
exc	str	The exchange.	required
plot	bool	Plot the prices. Defaults to True.	True
ax	Axes	The axes to plot on. Defaults to None.	None
with_fill	bool	Plot fills on the price plot. Defaults to True.	True
with_inventory	bool	Plot inventory on the price plot. Defaults to truth value of group is None.	None
key	str	If group is provided, the values of corresponding key are overlayed on the price plot. Defaults to None.	None
group	str	The group. Defaults to None.	None
share_index	bool	Share the x-axis index with the price plot for plotted key. Defaults to False.	False
show	bool	Show the plot. Defaults to True.	True
save	str	Saves the plot to specified filename if provided. Defaults to None.	None

get_feed()

Get a MockFeed instance that can be used like a quantpylib.hft.feed.Feed instance.

get_mid(exc, ticker)

Get the mid price known at current replay time.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required
ticker	str	The ticker.	required

get_oms()

Get a MockOMS instance that can be used like a quantpylib.hft.oms.OMS instance.

play() async

Play the replayer. To be used in place of await asyncio.sleep(x) that sleeps while the trading logic runs.

statistics()

Get a dictionary of statistics for each exchange after the replay.

statistics_exc(exc)

Get a dictionary of statistics for a specific exchange after the replay.

Parameters:

Name	Type	Description	Default
exc	str	The exchange.	required

time()

Get millisecond timestamp of the replayer is at during play.