Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Expire Instead of Clear Insights In Base PCM In OnSecuritiesChanged
* Fixes Regression Test
The regression was incorrect because `_removedSymbols` didn't retain the removed symbols from previous `OnSecuritiesChanged` calls, and the algorithm relances once per month. E.g. On day 1, SPY and FB were removed, on day 3 AAPL and IBM were removed and `removedSymbols` would only include AAPL and IBM.
We should hve fixed this problem before with:
```csharp
if (_removedSymbols == null) _removedSymbols = new List<Symbol>();
_removedSymbols.AddRange(changes.RemovedSecurities.Select(x => x.Symbol));
```
However the change to use Expire fixes the issue.
* Removes Unused List of Removed Symbols
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Add forward data only fill model example
- Add forward data only custom fill model C# & PY example.
- Minor adjustment for 'MarketOrderFillTimeout' to be zero always in
backtesting
* Address reviews
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Implement ShortableProviderPythonWrapper.cs
- Modify AllShortableSymbolsCoarseSelectionRegressionAlgorithm.cs and ShortableProviderOrdersRejectedRegressionAlgorithm.cs to use ShortableProvider from Security and not from the Brokerage model
- Add SetShortableProvider() overload method in Security.cs to set a custom shortable provider from Python
- Remove AllShortableSymbols() method from LocalDiskShortableProvider.cs
- Remove DefaultShortableProvider class
- Add regresion algorithms in C# to cover the changes done
* Implement ShortableProviderPythonWrapper.cs
- Modify AllShortableSymbolsCoarseSelectionRegressionAlgorithm.cs and ShortableProviderOrdersRejectedRegressionAlgorithm.cs to use ShortableProvider from Security and not from the Brokerage model
- Add SetShortableProvider() overload method in Security.cs to set a custom shortable provider from Python
- Remove AllShortableSymbols() method from LocalDiskShortableProvider.cs
- Remove DefaultShortableProvider class
- Add regresion algorithms in C# to cover the changes done
* Solve bugs and nit change
* Address review
---------
Co-authored-by: Martin-Molinero <martin@quantconnect.com>
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Add dataMappingMode parameter to every history api method overload
* Minor unit tests fixes
* Update regression algorithm stats
* Minor changes
* Minor changes
* Add extendedMarket parameter to every history api method overload
* Rename extendedMarketHours parameter
New name is extendedHours as in the History API to standarize parameters naming
* Update generic history overloads to use every matching subscription
* Update regression algorithms stats
* Centralize period-based history error for tick resolution
* Rename extended market hours parameter to extendedMarketHours
* Minor changes
* Minor changes
* Minor unit tests changes
* Minor unit tests changes
* Minor changes
* Minor unit tests changes
* Minor unit tests changes
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Add fillForward parameters to every History method
* Remove conflicting Python history method
* Undo removing conflicting Python history method
* Minor changes
* Minor changes
* Minor changes
* Add fillForward and extendedMarket parameters to history request factory
* Minor changes
* Minor changes
* Minor changes
* Minor changes
* Minor unit tests changes
* Rename fillForward parameter in History API
New name if fillDataForward as in the Add*Security API to standarize
parameters naming
* Rename fillForward parameter
Using the shorter fillForward in every API
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Add Collective2SignalExportClass
Add SignalExportTarget interface
* Collective2SignalExport test working
Add SignalExportManager
Add draft of CrunchDAOSignalExport
* Modify SignalExportManager
Instantiate SignalExportManager in QCAlgorithm constructor
Draft of CrunchDAOSignalExport
* Improve SignalExportManager
- Add regression tests SignalExportDemonstrationAlgorithm in C# and
Python
* Improve SignalExportDemonstrationAlgorithm
Address requested changes in Collective2SignalExport, SignalExportManger and SignalExportTargetTests.cs
* Add CrunchDAOSignalExport.cs
Add CrunchDAOSignalExport unit tests in SignalExportTargetTests.cs
* Add NumeraiSignalExport.cs
Modify SignalExportDemonstrationAlgorithm.cs to test NumeraiSignalExport
Add unit test in SignalExportTargetTests to test NumeraiSignalExport
* Address required changes
- Add BaseSignalExport.cs
- Add SignalExportParameters.cs
- Add PortfolioSignalExportDemonstrationAlgorithm.cs/py
- Improve Error handling in SignalExport provider classes
- Collective2SignalExport now gets the correct number of shares for each holding
- SignalExportManager now computes the correct holding percentage of each portfolio target
- SignalExportManager now takes into account if the algorithm is in live mode
- Demonstration algorithms now are more simple
* Address last required changes
- PortfolioSignalExportDemonstrationAlgorithm.cs/py now inherits from SignalExportDemonstrationAlgorithm.cs/py
- Add more unit tests to assert SignalExportManager gets the correct percentage quantity for each holding
- Change Collective2SignalExport, CrunchDAOSignalExport and NumeraiSignalExport Send() method to return true if there was no error while sending the signals and false otherwise
- Nit changes
* Remove exceptions thrown
- Add more unit tests and more test cases
- Enhance BaseSignalExport.Dispose() method
* Add Collective2SignalExportClass
Add SignalExportTarget interface
* Collective2SignalExport test working
Add SignalExportManager
Add draft of CrunchDAOSignalExport
* Modify SignalExportManager
Instantiate SignalExportManager in QCAlgorithm constructor
Draft of CrunchDAOSignalExport
* Improve SignalExportManager
- Add regression tests SignalExportDemonstrationAlgorithm in C# and
Python
* Improve SignalExportDemonstrationAlgorithm
Address requested changes in Collective2SignalExport, SignalExportManger and SignalExportTargetTests.cs
* Add CrunchDAOSignalExport.cs
Add CrunchDAOSignalExport unit tests in SignalExportTargetTests.cs
* Add NumeraiSignalExport.cs
Modify SignalExportDemonstrationAlgorithm.cs to test NumeraiSignalExport
Add unit test in SignalExportTargetTests to test NumeraiSignalExport
* Address required changes
- Add BaseSignalExport.cs
- Add SignalExportParameters.cs
- Add PortfolioSignalExportDemonstrationAlgorithm.cs/py
- Improve Error handling in SignalExport provider classes
- Collective2SignalExport now gets the correct number of shares for each holding
- SignalExportManager now computes the correct holding percentage of each portfolio target
- SignalExportManager now takes into account if the algorithm is in live mode
- Demonstration algorithms now are more simple
* Address last required changes
- PortfolioSignalExportDemonstrationAlgorithm.cs/py now inherits from SignalExportDemonstrationAlgorithm.cs/py
- Add more unit tests to assert SignalExportManager gets the correct percentage quantity for each holding
- Change Collective2SignalExport, CrunchDAOSignalExport and NumeraiSignalExport Send() method to return true if there was no error while sending the signals and false otherwise
- Nit changes
* Remove exceptions thrown
- Add more unit tests and more test cases
- Enhance BaseSignalExport.Dispose() method
* Fix failing regression tests
* Fix failing unit tests
* Nit changes
* Nit change
* Nit change
* Fix failing unit tests
* Changes required
- Break regression algos `SignalExportDemonstrationAlgorithm.cs/py` nad `PortfolioSignalExportDemonstrationAlgorithm.cs/`y` into three ones, one for each signal export provider
- Change SignalExportManager constructor to receive current algorithm as a parameter
- Fix bug in `SignalExportManager.GetPortfolioTargets()`, now it computes the correct percentage for each holding
- Make `BaseSignalExport.DefaultAllowedSecurityTypes` overrdible
- Handle case were `Collective2SignalExport.ConvertPercentageToQuantity()` returns null
- Clean unnecessary code in `Collective2SignalExport()`, `CrunchDAOSignalExport()` and `NumeraiSignalExport()`
* Nit change
* Nit change
* Minor tweaks after review
* Remove indexes from signal exports
* Required changes
- Change EMA indicators period from 200, 300 to 10,100 in regression algorithms
- Remove Indices from regression algorithms
- Add more XML documentation to regression algorithms
- Change `Log.Error` to `_algorithm.Error` in Signal export providers. Besides, fix error message format
- Change default value for `platformId` parameter in `Collective2SignalExport.cs` constructor
- Solve small bugs in SignalExportProvider when verificating the amount of porfolio targets is greater than zero and each portfolio target is allowed
- Handle case when `PortfolioTarget.Percent()` returns null in `Collective2SignalExport.ConvertPercentageToQuantity()`
- Handle error format message from Collective2 API
- Check every ticker signal is between 0 and 1 (inclusive) in `CrunchDAOSignalExport.cs`
- Modifiy `NumeraiSignalExport.cs` constructor to take into account filename given in the arguments
- Fix small bug with the return value of `ConvertTargetsToNumerai()` method in `NumeraiSignalExport.cs`
- Modify `SignalExportManager.cs` to return true when the algorithm being ran is not in live mode
- Remove indices from CrunchDAO unit tests
* Enhance ´CrunchDAOSignalExport.cs´ implementation
---------
Co-authored-by: Martin-Molinero <martin@quantconnect.com>
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Fix fill quantity using group order quantity for combo orders
* Add Order's ComboQuantity property
* Add Order's ComboDirection property
* Minor changes and regression algorithms update
* Minor changes
* Update algorithms stats
* Minor changes and regression algorithms update
* Store the full quantity for each combo order leg in Quantity property instead of the ratio
* Minor changes and regression algorithms update
* A few fixes after pair programming
* Handle grouping position reduction
---------
Co-authored-by: Martin-Molinero <martin@quantconnect.com>
Python Virtual Environments / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Asserts Number of Insights In One PCM Regression Test
If the `EqualWeightingPortfolioConstructionModel` interacts with the `QCAlgorithm.Insights`, the number of elements in the collection should not the sum of emitted insights.
* Refactor Portfolio Construction Models to Use Insight Manager
`PortfolioConstructionModel` will use `QCAlgorithm.Insights" instead of class property `InsightCollection` to manage the insights. It no longer adds insights to the collection, but it removes them if they expire or the securities are removed from the universe.
Updates PCMs that were affected by the change.
* Updates Unit Tests
We need to add the insights to the insight manager before we call `PortfolioConstruction.CreateTargets`
Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Apply splits and dividends to volatility models
* Apply splits and dividends to volatility models using history requests
* Add new ScaleRaw data normalization mode
Handling the new mode in the price scale enumerator.
* DataNormalizationMode.ScaledRaw history requests
* Minor changes
* Minor changes
* Disable new normalization mode in AddSecurity methods and other minor changes
* Peer review
* Minor changes
* Peer review
* Minor changes
* Peer review
* Peer review
* Peer review
* Add scaled raw history regression algorithm
* Add more regression algorithms
* Add more regression algorithms
* Add Slice.TryGet unit tests
* Peer review
* Peer review
* Peer review
* Peer review
* Peer review
* Update algorithms stats
* Peer review
* Peer review
Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Refactor alpha statistics
- Refactor alpha statistics, cleaning up and simplifying no longer required calculations and scoring
- Adding new InsightEvaluator abstraction, adding C# & PY regression
algorithms
* Optimization backtest result json converter update
* Address reviews
- Remove IAlphaHandler, move insight storage responsability to IResultHandler
and centralizing insight collection on the QCAlgorithm.Insights to be
reused by the framework models
- Fix portfolio turnover single day backtests and duplicate time
sampling handling. Updating regression algorithms
* Add InsightCollection tests and minor fixes
* Adding more & improved tests
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Adds Regression Test
The new regression test asserts that the total trades is 2.
* Fixes Liquidate Existing Holdings Bug
Liquidate existing holdings before open new postions.
* Addresses Peer-Review
* Fix live trading chart series index
- Fix Series loosing their index when handled by the
LiveTradingResultHandler
- Some clean up for Series.cs
* Fix SeriesSampler loosing Series settings
* Custom data type history request in python
* Potential solutions
* Minor changes
* Use Slice.Get(Type) for getting python custom data history
* Minor changes
* Add unit tests
* Add unit tests
* Udpdate regression algorithms
* Peer review
* Peer review
* Add research regression tests
* Minor changes
* Minor changes
* Minor tweaks
---------
Co-authored-by: Martin-Molinero <martin@quantconnect.com>
Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Expand Index option support
- Adjust APIs so that the same underlying can be used
for different options, adding support for SPX weekly options. Adding
regression tests
* Fix IndexOption.IsStandard
* Add IndexOption test
* Feature combo orders
- Add support for combo orders
* Make fill model wait for all grouped orders to emit fills
* Add ComboFill to model multiple fills for combo orders
* Fill combo limit orders
Add some regression algorithms
* Add fill implementation for combo leg limit orders
* Add IFill as common interface for Fill and ComboFill
* Refactor combo orders removing IGroupOrder interface
Move the group order manager to the base Order class
* Update algorithms
* Handle combo order events atomically
* Refactor brokerage transaction event handler
* Refactor combo fill models
* Process fills in batch
* Combo orders fill model tests
* Combo leg limit orders algorithm
* Regression algorithms cleanup
* Fill and combo fill classes cleanup
* Housekeeping
* Refactor equity fill model to derive from base fill model
* Address review changes request
* Handling the new types of orders in the OrderJsonConverter
* Add regression algorithm to test combo orders update/cancel
* Add regression algorithm to test combo orders update/cancel
* Housekeeping
* Address review changes request
* Minor changes
* Security transaction handler method for setting order request id
* Extend public interface for placing combo orders
* Combo order tickets demo algorithm python version
* Tweaks and updates
* Minor fixes
* Minor changes
* Minor fixes
* Address reviews minor fixes
* Minor fixes
Co-authored-by: Martin-Molinero <martin@quantconnect.com>
Python Virtual Environments / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
* WIP
* Add base currency cash
* Symbol properties and data processing
* Add basic template algorithm
* Add hourly crypto future algorithm
* Minor fixes after live trading testing
* CoinApiDataQueueHandler CryptoFuture support
* Address reviews
* Fix regression algorithms after update
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
* Allows Market-On-Close Orders Outside Buffer Period
Market-On-Close orders can be submitted before and after the buffer period from 15:45 to 16:00 (Tested with TWS) meaning that we can submit MOC when the market is closed and, consequently, use daily resolution data.
* Adds Regression Test with Extended Market Hours
- Removes `nextMarketClose > Time` condition since it's unnecessary. If the algorithm Time is greater than the close of that day, `nextMarektClose` refers to the next day.
* Updates Unit Test
Updates `OrderQuantityConversionTest` because the MOC orders are submitted. They are placed at 7 pm and invalid before this pull request.
* Updates Summary of new Regression Tests
* Fixes Regression Algorithm Testing No Warm Up Requirement
* Removes Volatility Model Warm Up Requirement
If the volatiliy model is not ready, we will use the Brenner and Subrahmanyam (1988) approximation (BS88).
Lean will not set the volatility model if the user doesn't, since the BS88 can handle the null model case.
* Memory Related Performance improvements
- Make sure we cleanup & dipose of python related objects during pandas
data generation.
- Disable memoizing enumerable use while creating pandas data frames,
since we do not require it
- Reduce unrequired object creations
- Replace concurrentCollections for ordinary locks
* Decimal parsing typo fix
Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
* Normalize time in universe behavior
- Normalize Option & Future chain universe behavior regarding their
assets time in universe. They will now respect the universe settings
time in universe value. Adding new regression algorithms asserting the
behavior
* Address reviews & cleanup
Regression Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Use `SetEndDate` to set `EndDate` in `QCAlgorithm` constructor. It will ensure that the `EndDate` is independent of the time the algorithm is executed if `SetEndDate` is not called in `Initilaize`.
Some users don't implement `SetEndDate` to run the algorithm to the latest datapoint, but this is not true if we run the algorithm during the day as the latest datapoint will be 24 hours before the execution time while there is data until the current day midnight.
Alexandre Catarino