Supported Indicators

Auto Regressive Integrated Moving Average

Introduction

An Autoregressive Intergrated Moving Average (ARIMA) is a time series model which can be used to describe a set of data. In particular,with Xₜ representing the series, the model assumes the data are of form (after differencing times): Xₜ = c + εₜ + ΣᵢφᵢXₜ₋ᵢ + Σᵢθᵢεₜ₋ᵢ where the first sum has an upper limit of and the second .

To view the implementation of this indicator, see the LEAN GitHub repository.

Using ARIMA Indicator

To create an automatic indicators for AutoRegressiveIntegratedMovingAverage, call the ARIMA helper method from the QCAlgorithm class. The ARIMA method creates a AutoRegressiveIntegratedMovingAverage object, hooks it up for automatic updates, and returns it so you can used it in your algorithm. In most cases, you should call the helper method in the Initializeinitialize method.

public class AutoRegressiveIntegratedMovingAverageAlgorithm : QCAlgorithm
{
    private Symbol _symbol;
    private AutoRegressiveIntegratedMovingAverage _arima;

    public override void Initialize()
    {
        _symbol = AddEquity("SPY", Resolution.Daily).Symbol;
        _arima = ARIMA(_symbol, 1, 1, 1, 20);
    }

    public override void OnData(Slice data)
    {
        if (_arima.IsReady)
        {
            // The current value of _arima is represented by itself (_arima)
            // or _arima.Current.Value
            Plot("AutoRegressiveIntegratedMovingAverage", "arima", _arima);
            // Plot all properties of arima
            Plot("AutoRegressiveIntegratedMovingAverage", "handleexceptions", _arima.HandleExceptions);
            Plot("AutoRegressiveIntegratedMovingAverage", "arparameters", _arima.ArParameters);
            Plot("AutoRegressiveIntegratedMovingAverage", "maparameters", _arima.MaParameters);
            Plot("AutoRegressiveIntegratedMovingAverage", "intercept", _arima.Intercept);
            Plot("AutoRegressiveIntegratedMovingAverage", "arresidualerror", _arima.ArResidualError);
            Plot("AutoRegressiveIntegratedMovingAverage", "maresidualerror", _arima.MaResidualError);
        }
    }
}
class AutoRegressiveIntegratedMovingAverageAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol
        self._arima = self.arima(self._symbol, 1, 1, 1, 20)

    def on_data(self, slice: Slice) -> None:
        if self._arima.is_ready:
            # The current value of self._arima is represented by self._arima.current.value
            self.plot("AutoRegressiveIntegratedMovingAverage", "arima", self._arima.current.value)
            # Plot all attributes of self._arima
            self.plot("AutoRegressiveIntegratedMovingAverage", "handle_exceptions", self._arima.handle_exceptions)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ar_parameters", self._arima.ar_parameters)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ma_parameters", self._arima.ma_parameters)
            self.plot("AutoRegressiveIntegratedMovingAverage", "intercept", self._arima.intercept)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ar_residual_error", self._arima.ar_residual_error)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ma_residual_error", self._arima.ma_residual_error)

The following reference table describes the ARIMA method:

arima(symbol, ar_order, diff_order, ma_order, period, resolution=None, selector=None)[source]

Creates a new ARIMA indicator.

Parameters:
  • symbol (Symbol) — The symbol whose ARIMA indicator we want
  • ar_order (int) — AR order (p) -- defines the number of past values to consider in the AR component of the model.
  • diff_order (int) — Difference order (d) -- defines how many times to difference the model before fitting parameters.
  • ma_order (int) — MA order (q) -- defines the number of past values to consider in the MA component of the model.
  • period (int) — Size of the rolling series to fit onto
  • resolution (Resolution, optional) — The resolution
  • selector (Callable[IBaseData, float], optional) — x.Value)
Returns:

The ARIMA indicator for the requested symbol over the specified period

Return type:

AutoRegressiveIntegratedMovingAverage

ARIMA(symbol, arOrder, diffOrder, maOrder, period, resolution=None, selector=None)[source]

Creates a new ARIMA indicator.

Parameters:
  • symbol (Symbol) — The symbol whose ARIMA indicator we want
  • arOrder (Int32) — AR order (p) -- defines the number of past values to consider in the AR component of the model.
  • diffOrder (Int32) — Difference order (d) -- defines how many times to difference the model before fitting parameters.
  • maOrder (Int32) — MA order (q) -- defines the number of past values to consider in the MA component of the model.
  • period (Int32) — Size of the rolling series to fit onto
  • resolution (Resolution, optional) — The resolution
  • selector (Func<IBaseData, Decimal>, optional) — x.Value)
Returns:

The ARIMA indicator for the requested symbol over the specified period

Return type:

AutoRegressiveIntegratedMovingAverage

If you don't provide a resolution, it defaults to the security resolution. If you provide a resolution, it must be greater than or equal to the resolution of the security. For instance, if you subscribe to hourly data for a security, you should update its indicator with data that spans 1 hour or longer.

For more information about the selector argument, see Alternative Price Fields.

For more information about plotting indicators, see Plotting Indicators.

You can manually create a AutoRegressiveIntegratedMovingAverage indicator, so it doesn't automatically update. Manual indicators let you update their values with any data you choose.

Updating your indicator manually enables you to control when the indicator is updated and what data you use to update it. To manually update the indicator, call the Updateupdate method with time/number pair or an IndicatorDataPoint. The indicator will only be ready after you prime it with enough data.

public class AutoRegressiveIntegratedMovingAverageAlgorithm : QCAlgorithm
{
    private Symbol _symbol;
    private AutoRegressiveIntegratedMovingAverage _arima;

    public override void Initialize()
    {
        _symbol = AddEquity("SPY", Resolution.Daily).Symbol;
        _arima = new AutoRegressiveIntegratedMovingAverage(1, 1, 1, 20, True);
    }

    public override void OnData(Slice data)
    {
        if (data.Bars.TryGetValue(_symbol, out var bar))
        {      
            _arima.Update(bar.EndTime, bar.Close);
        }
   
        if (_arima.IsReady)
        {
            // The current value of _arima is represented by itself (_arima)
            // or _arima.Current.Value
            Plot("AutoRegressiveIntegratedMovingAverage", "arima", _arima);
            // Plot all properties of arima
            Plot("AutoRegressiveIntegratedMovingAverage", "handleexceptions", _arima.HandleExceptions);
            Plot("AutoRegressiveIntegratedMovingAverage", "arparameters", _arima.ArParameters);
            Plot("AutoRegressiveIntegratedMovingAverage", "maparameters", _arima.MaParameters);
            Plot("AutoRegressiveIntegratedMovingAverage", "intercept", _arima.Intercept);
            Plot("AutoRegressiveIntegratedMovingAverage", "arresidualerror", _arima.ArResidualError);
            Plot("AutoRegressiveIntegratedMovingAverage", "maresidualerror", _arima.MaResidualError);
        }
    }
}
class AutoRegressiveIntegratedMovingAverageAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol
        self._arima = AutoRegressiveIntegratedMovingAverage(1, 1, 1, 20, True)

    def on_data(self, slice: Slice) -> None:
        bar = slice.bars.get(self._symbol)
        if bar:
            self._arima.update(bar.EndTime, bar.Close)
        if self._arima.is_ready:
            # The current value of self._arima is represented by self._arima.current.value
            self.plot("AutoRegressiveIntegratedMovingAverage", "arima", self._arima.current.value)
            # Plot all attributes of self._arima
            self.plot("AutoRegressiveIntegratedMovingAverage", "handle_exceptions", self._arima.handle_exceptions)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ar_parameters", self._arima.ar_parameters)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ma_parameters", self._arima.ma_parameters)
            self.plot("AutoRegressiveIntegratedMovingAverage", "intercept", self._arima.intercept)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ar_residual_error", self._arima.ar_residual_error)
            self.plot("AutoRegressiveIntegratedMovingAverage", "ma_residual_error", self._arima.ma_residual_error)

To register a manual indicator for automatic updates with the security data, call the RegisterIndicatorregister_indicator method.

public class AutoRegressiveIntegratedMovingAverageAlgorithm : QCAlgorithm
{
    private Symbol _symbol;
    private AutoRegressiveIntegratedMovingAverage _arima;

    public override void Initialize()
    {
        _symbol = AddEquity("SPY", Resolution.Daily).Symbol;
        _arima = new AutoRegressiveIntegratedMovingAverage(1, 1, 1, 20, True);
        RegisterIndicator(_symbol, _arima, Resolution.Daily);
    }

    public override void OnData(Slice data)
    {
        if (_arima.IsReady)
        {
            // The current value of _arima is represented by itself (_arima)
            // or _arima.Current.Value
            Plot("AutoRegressiveIntegratedMovingAverage", "arima", _arima);
            
        }
    }
}
class AutoRegressiveIntegratedMovingAverageAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol
        self._arima = AutoRegressiveIntegratedMovingAverage(1, 1, 1, 20, True)
        self.register_indicator(self._symbol, self._arima, Resolution.DAILY)

    def on_data(self, slice: Slice) -> None:
        if self._arima.is_ready:
            # The current value of self._arima is represented by self._arima.current.value
            self.plot("AutoRegressiveIntegratedMovingAverage", "arima", self._arima.current.value)

The following reference table describes the AutoRegressiveIntegratedMovingAverage constructor:

AutoRegressiveIntegratedMovingAverage

class QuantConnect.Indicators.AutoRegressiveIntegratedMovingAverage[source]

An Autoregressive Intergrated Moving Average (ARIMA) is a time series model which can be used to describe a set of data. In particular,with Xₜ representing the series, the model assumes the data are of form (after differencing _diffOrder times): Xₜ = c + εₜ + ΣᵢφᵢXₜ₋ᵢ + Σᵢθᵢεₜ₋ᵢ where the first sum has an upper limit of _arOrder and the second _maOrder.

get_enumerator()

Returns an enumerator that iterates through the history window.

Return type:

IEnumerator[IndicatorDataPoint]

reset()

Resets this indicator to its initial state

to_detailed_string()

Provides a more detailed string of this indicator in the form of {Name} - {Value}

Return type:

str

update(time, value)

Updates the state of this indicator with the given value and returns true if this indicator is ready, false otherwise

Parameters:
  • time (datetime)
  • value (float)
Return type:

bool

update(input)

Updates the state of this indicator with the given value and returns true if this indicator is ready, false otherwise

Parameters:
  • input (IBaseData)
Return type:

bool

property ar_parameters

Fitted AR parameters (φ terms).

Returns:

Fitted AR parameters (φ terms).

Return type:

float[]

property ar_residual_error

The variance of the residuals (Var(ε)) from the first step of Double[]).

Returns:

The variance of the residuals (Var(ε)) from the first step of Double[]).

Return type:

float

property consolidators

The data consolidators associated with this indicator if any

Returns:

The data consolidators associated with this indicator if any

Return type:

ISet[IDataConsolidator]

property current

Gets the current state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Returns:

Gets the current state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Return type:

IndicatorDataPoint

property handle_exceptions

Whether or not to handle potential exceptions, returning a zero value. I.e, the values provided as input are not valid by the Normal Equations direct regression method

Returns:

Whether or not to handle potential exceptions, returning a zero value. I.e, the values provided as input are not valid by the Normal Equations direct regression method

Return type:

bool

property intercept

Fitted intercept (c term).

Returns:

Fitted intercept (c term).

Return type:

float

property is_ready

Gets a flag indicating when this indicator is ready and fully initialized

Returns:

Gets a flag indicating when this indicator is ready and fully initialized

Return type:

bool

property item

Indexes the history windows, where index 0 is the most recent indicator value. If index is greater or equal than the current count, it returns null. If the index is greater or equal than the window size, it returns null and resizes the windows to i + 1.

Returns:

Indexes the history windows, where index 0 is the most recent indicator value. If index is greater or equal than the current count, it returns null. If the index is greater or equal than the window size, it returns null and resizes the windows to i + 1.

Return type:

IndicatorDataPoint

property ma_parameters

Fitted MA parameters (θ terms).

Returns:

Fitted MA parameters (θ terms).

Return type:

float[]

property ma_residual_error

The variance of the residuals (Var(ε)) from the second step of Double[]).

Returns:

The variance of the residuals (Var(ε)) from the second step of Double[]).

Return type:

float

property name

Gets a name for this indicator

Returns:

Gets a name for this indicator

Return type:

str

property previous

Gets the previous state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Returns:

Gets the previous state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Return type:

IndicatorDataPoint

property samples

Gets the number of samples processed by this indicator

Returns:

Gets the number of samples processed by this indicator

Return type:

int

property warm_up_period

Required period, in data points, for the indicator to be ready and fully initialized.

Returns:

Required period, in data points, for the indicator to be ready and fully initialized.

Return type:

int

property window

A rolling window keeping a history of the indicator values of a given period

Returns:

A rolling window keeping a history of the indicator values of a given period

Return type:

RollingWindow[IndicatorDataPoint]

AutoRegressiveIntegratedMovingAverage

class QuantConnect.Indicators.AutoRegressiveIntegratedMovingAverage[source]

An Autoregressive Intergrated Moving Average (ARIMA) is a time series model which can be used to describe a set of data. In particular,with Xₜ representing the series, the model assumes the data are of form (after differencing _diffOrder times): Xₜ = c + εₜ + ΣᵢφᵢXₜ₋ᵢ + Σᵢθᵢεₜ₋ᵢ where the first sum has an upper limit of _arOrder and the second _maOrder.

GetEnumerator()

Returns an enumerator that iterates through the history window.

Return type:

IEnumerator[IndicatorDataPoint]

Reset()

Resets this indicator to its initial state

ToDetailedString()

Provides a more detailed string of this indicator in the form of {Name} - {Value}

Return type:

String

Update(time, value)

Updates the state of this indicator with the given value and returns true if this indicator is ready, false otherwise

Parameters:
  • time (DateTime)
  • value (decimal)
Return type:

Boolean

Update(input)

Updates the state of this indicator with the given value and returns true if this indicator is ready, false otherwise

Parameters:
  • input (IBaseData)
Return type:

Boolean

property ArParameters

Fitted AR parameters (φ terms).

Returns:

Fitted AR parameters (φ terms).

Return type:

Double[]

property ArResidualError

The variance of the residuals (Var(ε)) from the first step of Double[]).

Returns:

The variance of the residuals (Var(ε)) from the first step of Double[]).

Return type:

Double

property Consolidators

The data consolidators associated with this indicator if any

Returns:

The data consolidators associated with this indicator if any

Return type:

ISet<IDataConsolidator>

property Current

Gets the current state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Returns:

Gets the current state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Return type:

IndicatorDataPoint

property HandleExceptions

Whether or not to handle potential exceptions, returning a zero value. I.e, the values provided as input are not valid by the Normal Equations direct regression method

Returns:

Whether or not to handle potential exceptions, returning a zero value. I.e, the values provided as input are not valid by the Normal Equations direct regression method

Return type:

bool

property Intercept

Fitted intercept (c term).

Returns:

Fitted intercept (c term).

Return type:

Double

property IsReady

Gets a flag indicating when this indicator is ready and fully initialized

Returns:

Gets a flag indicating when this indicator is ready and fully initialized

Return type:

bool

property MaParameters

Fitted MA parameters (θ terms).

Returns:

Fitted MA parameters (θ terms).

Return type:

Double[]

property MaResidualError

The variance of the residuals (Var(ε)) from the second step of Double[]).

Returns:

The variance of the residuals (Var(ε)) from the second step of Double[]).

Return type:

Double

property Name

Gets a name for this indicator

Returns:

Gets a name for this indicator

Return type:

string

property Previous

Gets the previous state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Returns:

Gets the previous state of this indicator. If the state has not been updated then the time on the value will equal DateTime.MinValue.

Return type:

IndicatorDataPoint

property Samples

Gets the number of samples processed by this indicator

Returns:

Gets the number of samples processed by this indicator

Return type:

int

property WarmUpPeriod

Required period, in data points, for the indicator to be ready and fully initialized.

Returns:

Required period, in data points, for the indicator to be ready and fully initialized.

Return type:

Int32

property Window

A rolling window keeping a history of the indicator values of a given period

Returns:

A rolling window keeping a history of the indicator values of a given period

Return type:

RollingWindow<IndicatorDataPoint>

property [System.Int32]

Indexes the history windows, where index 0 is the most recent indicator value. If index is greater or equal than the current count, it returns null. If the index is greater or equal than the window size, it returns null and resizes the windows to i + 1.

Returns:

Indexes the history windows, where index 0 is the most recent indicator value. If index is greater or equal than the current count, it returns null. If the index is greater or equal than the window size, it returns null and resizes the windows to i + 1.

Return type:

IndicatorDataPoint

Visualization

The following image shows plot values of selected properties of AutoRegressiveIntegratedMovingAverage using the plotly library.

AutoRegressiveIntegratedMovingAverage line plot.

Indicator History

To get the historical data of the AutoRegressiveIntegratedMovingAverage indicator, call the IndicatorHistoryself.indicator_history method. This method resets your indicator, makes a history request, and updates the indicator with the historical data. Just like with regular history requests, the IndicatorHistoryindicator_history method supports time periods based on a trailing number of bars, a trailing period of time, or a defined period of time. If you don't provide a resolution argument, it defaults to match the resolution of the security subscription. 

public class AutoRegressiveIntegratedMovingAverageAlgorithm : QCAlgorithm
{
    private Symbol _symbol;

    public override void Initialize()
    {
        _symbol = AddEquity("SPY", Resolution.Daily).Symbol;
        var arima = ARIMA(_symbol, 1, 1, 1, 20);
        var countIndicatorHistory = IndicatorHistory(arima, _symbol, 100, Resolution.Minute);
        var timeSpanIndicatorHistory = IndicatorHistory(arima, _symbol, TimeSpan.FromDays(10), Resolution.Minute);
        var timePeriodIndicatorHistory = IndicatorHistory(arima, _symbol, new DateTime(2024, 7, 1), new DateTime(2024, 7, 5), Resolution.Minute);
    }
}
class AutoRegressiveIntegratedMovingAverageAlgorithm(QCAlgorithm):
    def initialize(self) -> None:
        self._symbol = self.add_equity("SPY", Resolution.DAILY).symbol
        arima = self.arima(self._symbol, 1, 1, 1, 20)
        count_indicator_history = self.indicator_history(arima, self._symbol, 100, Resolution.MINUTE)
        timedelta_indicator_history = self.indicator_history(arima, self._symbol, timedelta(days=10), Resolution.MINUTE)
        time_period_indicator_history = self.indicator_history(arima, self._symbol, datetime(2024, 7, 1), datetime(2024, 7, 5), Resolution.MINUTE)

To make the IndicatorHistoryindicator_history method update the indicator with an alternative price field instead of the close (or mid-price) of each bar, pass a selector argument.

var indicatorHistory = IndicatorHistory(arima, 100, Resolution.Minute, (bar) => ((TradeBar)bar).High);
indicator_history = self.indicator_history(arima, 100, Resolution.MINUTE, lambda bar: bar.high)
indicator_history_df = indicator_history.data_frame

If you already have a list of Slice objects, you can pass them to the IndicatorHistoryindicator_history method to avoid the internal history request.

var history = History(_symbol, 100, Resolution.Minute);
var historyIndicatorHistory = IndicatorHistory(arima, history);

You can also see our Videos. You can also get in touch with us via Discord.

Did you find this page helpful?

Contribute to the documentation: