Zipline
Quick Reference
User-Implemented Functions
The following table describes functions that are treated specially when defined in the IDE.
Algorithms are required to implement one method: Initialize(self).
An optional scheduled event can be defined to behave like before_trading_start(). OnData() is also optional but is treated especially if defined.
| Quantopian | QuantConnect |
|---|---|
initialize(context) | Initialize(self) |
| Required function called once at the start of a backtest. | |
before_trading_start(context, data) | self.Schedule.On(self.DateRules.EveryDay(), self.TimeRules.AfterMarketOpen(symbol, 10), self.before_trading_start) |
| Optional function called prior to market open on every trading day. | |
handle_data(context, data) | OnData(self, slice) |
| Optional function called at the end of each market minute. | |
Interface Classes
The following table describes classes that are passed to initialize(), handle_data(), before_trading_start() and any scheduled with schedule_function().
In QuantConnect/Lean, the portfolio state and the positions are class attributes of QCAlgorithm, available via self keyword. The scheduled function is parameterless and can algorithms can self.CurrentSlice.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.interface.AlgorithmContext | QCAlgorithm |
| Shared object for storing custom state. | |
quantopian.algorithm.interface.BarData | Slice |
| Provides methods for accessing minutely and daily price/volume data from Algorithm API functions. | |
quantopian.algorithm.interface.Portfolio | self.Portfolio |
| Object providing read-only access to current portfolio state. | |
quantopian.algorithm.interface.Positions | self.Transactions |
| A dict-like object containing the algorithm's current positions. | |
The features from the BarData methods are found in the Slice object (current data), History method (past data) and class attribute Securities.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.interface.BarData.current | Slice[symbol] |
| Returns the "current" value of the given fields for the given assets at the current simulation time. | |
quantopian.algorithm.interface.BarData.history | self.History(symbols, bar_count) |
Returns a trailing window of length bar_count containing data for the given assets, fields, and frequency. | |
quantopian.algorithm.interface.BarData.can_trade | self.Securities[symbol].IsTradable |
| For the given asset or iterable of assets, returns True if all of the following are true: | |
quantopian.algorithm.interface.BarData.is_stale | not Slice[symbol].IsFillForward |
| For the given asset or iterable of assets, returns True if the asset is alive and there is no trade data for the current simulation time. | |
Pipeline
The following table describes functions used to schedule and retrieve results of Universe selection. For more information on the Universe Selection API, see the Univeres API Reference.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.attach_pipeline(...) | universe = self.AddUniverse(...)name = universe.Configuration.Symbol |
| Register a pipeline to be computed at the start of each day. | |
quantopian.algorithm.pipeline_output(name) | self.UniverseManager[name] |
Get results of the pipeline attached by with name. | |
Scheduling Functions
The following tables describes functions that can be used to schedule functions to run periodically during your algorithm.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.schedule_function(func) | self.Schedule.On(IDateRule, ITimeRule, func) |
| Schedule a function to be called repeatedly in the future. | |
Date Rules
The following table contains functions that can be used with the date_rule parameter of schedule_function().
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.date_rules.every_day() | self.DateRules.EveryDay(symbol) |
| Create a rule that triggers every day. | |
quantopian.algorithm.date_rules.month_start([...]) | self.DateRules.MonthStart(symbol, daysOffset) |
| Create a rule that triggers a fixed number of trading days after the start of each month. | |
quantopian.algorithm.date_rules.month_end([...]) | self.DateRules.MonthEnd(symbol, daysOffset) |
| Create a rule that triggers a fixed number of trading days before the end of each month. | |
quantopian.algorithm.date_rules.week_start([...]) | self.DateRules.WeekStart(symbol, daysOffset) |
| Create a rule that triggers a fixed number of trading days after the start of each week. | |
quantopian.algorithm.date_rules.week_end([...]) | self.DateRules.WeekEnd(symbol, daysOffset) |
| Create a rule that triggers a fixed number of trading days before the end of each week. | |
Time Rules
The following table contains functions that can be used with the time_rule parameter of schedule_function().
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.time_rules.market_open([...]) | self.TimeRules.AfterMarketOpen(symbol, minutes_after_open = 0) |
| Create a rule that triggers at a fixed offset from market open. | |
quantopian.algorithm.time_rules.market_close([...]) | self.TimeRules.BeforeMarketClose(symbol, minutes_before_close = 0) |
| Create a rule that triggers at a fixed offset from market close. | |
Ordering
There are several ways to place orders from an algorithm. For most use-cases, we recommend using SetHoldings method which correctly accounts the porfolio buying power model and fees. At the moment, there is no equivalent for order_optimal_portfolio(). However a similar behavior can be replicated with a Portfolio Construction Model.
Algorithms that require explicit manual control of their orders can use the lower-level ordering functions.
Placing Orders
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.order_optimal_portfolio(...) | N/A |
| Calculate an optimal portfolio and place orders toward that portfolio. | |
quantopian.algorithm.order(asset, amount[, ...]) | self.Order(symbol, amount) |
| Place an order for a fixed number of shares. | |
quantopian.algorithm.order_value(asset, value) | N/A |
| Place an order for a fixed amount of money. | |
quantopian.algorithm.order_percent(asset, ...) | self.SetHoldings(symbol, percent) |
| Place an order in the specified asset corresponding to the given percent of the current portfolio value. | |
quantopian.algorithm.order_target(asset, target) | N/A |
| Place an order to adjust a position to a target number of shares. | |
quantopian.algorithm.order_target_value(...) | N/A |
| Place an order to adjust a position to a target value. | |
quantopian.algorithm.order_target_percent(...) | self.SetHoldings(symbol, percent) |
| Place an order to adjust a position to a target percent of the current portfolio value. | |
Controlling Order Execution
| Quantopian | QuantConnect |
|---|---|
zipline.finance.execution.ExecutionStyle | FillModel |
| Base class for order execution styles. | |
zipline.finance.execution.MarketOrder([exchange]) | self.MarketOrder(symbol, quantity) |
| Execution style for orders to be filled at current market price. | |
zipline.finance.execution.LimitOrder(limit_price) | self.LimitOrder(symbol, quantity, limit_price) |
| Execution style for orders to be filled at a price equal to or better than a specified limit price. | |
zipline.finance.execution.StopOrder(stop_price) | self.StopMarketOrder(symbol, quantity, stop_price) |
| Execution style representing a market order to be placed if market price reaches a threshold. | |
zipline.finance.execution.StopLimitOrder(...) | self.StopLimitOrder(symbol, quantity, stop_price, limit_price) |
| Execution style representing a limit order to be placed if market price reaches a threshold. | |
Managing Existing Orders
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.get_order(order_id) | self.Transactions.GetOrderById(order_id) |
| Lookup an order based on the order Id returned from one of the order functions. | |
quantopian.algorithm.get_open_orders([asset]) | self.Transactions.GetOpenOrders(symbol) |
| Retrieve all of the current open orders. | |
quantopian.algorithm.cancel_order(order_param) | self.Transactions.CancelOpenOrders(symbol) |
| Cancel an open order. | |
Customizing the Simulation
Slippage
Algorithms can customize the simulation of order fills by calling SetSlippageModel() with a ISlippageModel which is any class that implements GetSlippageApproximation method. Unlike Quantopian, the slippage model is a security attribute, not a global attribute. Different securities can have different slippage models.
The default slippage model on QuantConnect is a constant slippage model with zero slippage percent (no slippage).
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.set_slippage([...]) | self.Securities[symbol].SetSlippageModel(ISlippageModel) |
| Set the slippage models for the simulation. | |
zipline.finance.slippage.SlippageModel() | N/A (class must implement GetSlippageApproximation method) |
| Abstract base class for slippage models. | |
zipline.finance.slippage.FixedBasisPointsSlippage([...]) | ConstantSlippageModel(slippage_percent) |
| Model slippage as a fixed percentage difference from historical minutely close price, limiting the size of fills to a fixed percentage of historical minutely volume. | |
zipline.finance.slippage.NoSlippage() | NullSlippageModel.Instance |
| A slippage model where all orders fill immediately and completely at the current close price. | |
zipline.finance.slippage.FixedSlippage([spread]) | N/A |
| Simple model assuming a fixed-size spread for all assets. | |
zipline.finance.slippage.VolumeShareSlippage([...]) | N/A |
| Model slippage as a quadratic function of percentage of historical volume. | |
Commissions
Algorithms can customize the simulation of order fees by calling SetFeeModel() with a IFeeModel which is any class that implements GetOrderFee method. Unlike Quantopian, the fee model is a security attribute, not a global attribute. Different securities can have different fee models.
The default fee model on QuantConnect depends on the security type and the brokerage. For instance, the default fee model for equity is the Interactive Brokers fee model.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.set_commission([...]) | self.Securities[symbol].SetFeeModel(IFeeModel) |
| Sets the commission models for the simulation. | |
zipline.finance.commission.CommissionModel | N/A (class must implement GetOrderFee method) |
| Abstract base class for commission models. | |
zipline.finance.commission.PerShare([cost, ...]) | N/A |
| Calculates a commission for a transaction based on a per share cost with an optional minimum cost per trade. | |
zipline.finance.commission.PerTrade([cost]) | N/A |
| Calculates a commission for a transaction based on a per trade cost. | |
zipline.finance.commission.PerDollar([cost]) | N/A |
| Model commissions by applying a fixed cost per dollar transacted. | |
zipline.finance.commission.NoCommission | FeeModel() |
| Model commissions as free. | |
Benchmark
The default benchmark is SPY.
| Quantopian | QuantConnect |
|---|---|
quantopian.algorithm.set_benchmark(benchmark) | self.SetBenchmark(benchmark) |
| Set the benchmark asset. | |
You can also see our Videos. You can also get in touch with us via Discord.
Did you find this page helpful?
