IlshatGaripov
29e9d678f2
* Fixes 4815 by loading the requested assembly from different folder. # Conflicts: # ToolBox/Program.cs * Upgrades System.Collections.Immutable to Version=1.2.5.0 * Creates a prototype for SSE streaming in IEXDataQueueHandler. * Revert the changes in Tick.cs * Implements a logic in IEXDataQueueHandler that updates the data-feed subscription after Subscribe/Unsubscribe * Implements IEXCouldSubscribeMoreThan100Symbols - which fails and other small fixes. * Implements DoForEach LinqExtensions * Implements IEXEventSourceCollection that wraps all logic that is SSE-subscriptions and symbol-limits-per-connection concerned. * Changes: 1) Fixes to address review. 2) Makes IexMarketPercent in QuoteSSE nullable as null values are assigned to in this field in data object received before the traing session start. 3) Deprecates helper Subscribe/Unsubscribe in IEXDataQueueHandler and IEXCouldSubscribe test. * Fixes: 1) _refreshEvent.Reset() order was not correct - should be called before UpdateSubscription 2) ProcessJsonObject- leaves only the functionality to emit ticks. 3) IEXEventSourceCollection - replaces int counter with CountdownEvent to improve the logic - in particular, need a mechanism that would not allow the repeated call to continue until the first one is completed * Refines the logic with parsing a data snapshot. * Fixes few more bugs: 1) Logic in ProcessJsonObject 2) Logic in UpdateSubscription - need to introduce additional ManualResetEvent to implement the intended logic - otherwise the logic is not suitable for general case * Introduce rate-gate limit in IEXEventSourceCollection: because when subscribing to a bunch of shares (more than 200 for example) the violation of rate gate policy may occur, which described in API docs asRequest Limits IEX Cloud only applies request limits per IP address to ensure system stability. We limit requests to 100 per second per IP measured in milliseconds, so no more than 1 request per 10 milliseconds. SSE endpoints are limited to 50 symbols per connection. You can make multiple connections if you need to consume more than 50 symbols.: * Few additional fixes done after real time testing * Adds xml-docs in stream response object + renaming a file. * Fixes: 1) Additional StreamResponseStocksUS parsing issues, that can happen outside of regular exchange hours. 2) Cancel clientUpdateThread by means of CancellationTokenSource 3) Replace BuildSymbolsQuery by string.Join * Fixes: 1) Changing Log Trace -> Debug 2) Adds ConfigureAwait(false) to async method call 3) Removes direct reference to System.Net.Http * Removes a task and manual reset event in IEXEventSourceCollection * Additions: 1) IEXEventSourceCollectionSubscribes test 2) GetSnpStocksArray() helper method 3) Installs packages in QC.tests : HtmlAgilityPack & LaunchDarkly.EventSource * IEX history provider fixes : 1) Tiny bug in ProcessJsonObject - use continue instead of return as execution is inside the for-each block) 2) Adds period variable for the historical data retrieved 3) Fixing from if (date.Date < start.Date || date.Date > end.Date) conditional check --> if (date < start || date > end) for more precise sorting. * Changes: 1) Removes HtmlAgilityPack and SNP scraper 2) Uses hard coded symbols instead * Bug fix: - at certain hours (example: before pre-market open or on holidays) IEX may send no data on subscription - when trying to connect during those hours Message handler may not be fired - need to place the counter signal to client.Opened to be informed of successful connect. * Implements: 1) IEXEventSourceCollectionSubscriptionThoroughTest and MockedIEXEventSourceCollection 2) Makes changes to IEXEventSourceCollection accordingly to allow the thorough testing. * Fixes formatting issue in StreamResponseStocksUS * Small fix for a new tests: - Change RemovedClientSymbols to keep not clients itself, but symbols array, because clients are being disposed right further * Enables extended logging in Toolbox. * Fixing IEX historical data fetcher bugs: 1) Bug in IEXDataDownloader.cs - HistoryRequest not precisely correct. 2) Enables day-by-day daily bar downloading in IEXDataQueueHandler. Motivation: Suppose we need data for some interval in the past - from-date=20170915-00:00:00 --to-date=20171103-00:00:00. With current behavior IEX would have to download all the historical data from =20170915-00:00:00 up to this day. 3) Extends SynchronizingHistoryProvider * Enables async fashion historical data download * More fixes to IEXCouldGetHistory test. * Reverts day-by-day daily bar downloading and other fixes. * Removes needless packages & references * Fix package reference * To address review * Sort out zero price ticks: after testing on real-time algo 30 min before the market open now - IEX may send updates for many securities with zero lastPrice, lastSize - fix to sort such entries out * Workaround for missing QuoteTicks timestamps: Since we don't have a stamp for quote tick updates (only for trades) we calculate the average delay between trade tick's time stamp and local time, and assuming that delay in average is the same for quote updates - just assign the local machine time adjusted for this average * Simplifies the things. * Changes: 1) Deprecates quote updates for IEX stocks. 2) Reduce the stream updates to reduce costs to -> # Stock Quotes every 1 second (per symbol? ) # Can be up to 54,000 messages per symbol per day https://iexcloud.io/docs/api/#how-messages-work * Fixes: 1) IEXDataQueueHandler: give an error message on extended market hours or tick resolution subscription request. As they are not really well supported by IEX. 2) Few small fixes in IEXEventSourceCollection, including additional condition for when the subscription remains irrevocable.
Lean Home - https://www.quantconnect.com/lean | Documentation | Download Zip | Docker Hub
Introduction
Lean Engine is an open-source algorithmic trading engine built for easy strategy research, backtesting and live trading. We integrate with common data providers and brokerages so you can quickly deploy algorithmic trading strategies.
The core of the LEAN Engine is written in C#; but it operates seamlessly on Linux, Mac and Windows operating systems. It supports algorithms written in Python 3.6, C# or F#. Lean drives the web-based algorithmic trading platform QuantConnect.
Proudly Sponsored By
Want your company logo here? Sponsor LEAN to be part of radically open algorithmic-trading innovation.
QuantConnect is Hiring!
Join the team and solve some of the most difficult challenges in quantitative finance. If you are passionate about algorithmic trading we'd like to hear from you. The below roles are open in our Seattle, WA office. When applying, make sure to mention you came through GitHub:
-
Senior UX Developer: Collaborate with QuantConnect to develop a world-leading online experience for a community of developers from all over the world.
-
Technical Writers: Help us improve the QuantConnect and LEAN documentation with hands-on tutorials with how to use all the adaptors LEAN offers, and how to set up trading locally.
-
Quantitative Development Intern: If you are a recent or current graduate with a knack for quantitative finance, consider applying for an internship!
System Overview
The Engine is broken into many modular pieces which can be extended without touching other files. The modules are configured in config.json as set "environments". Through these environments, you can control LEAN to operate in the mode required.
The most important plugins are:
-
Result Processing (IResultHandler)
Handle all messages from the algorithmic trading engine. Decide what should be sent, and where the messages should go. The result processing system can send messages to a local GUI, or the web interface.
-
Datafeed Sourcing (IDataFeed)
Connect and download the data required for the algorithmic trading engine. For backtesting this sources files from the disk, for live trading, it connects to a stream and generates the data objects.
-
Transaction Processing (ITransactionHandler)
Process new order requests; either using the fill models provided by the algorithm or with an actual brokerage. Send the processed orders back to the algorithm's portfolio to be filled.
-
Realtime Event Management (IRealtimeHandler)
Generate real-time events - such as the end of day events. Trigger callbacks to real-time event handlers. For backtesting, this is mocked-up a works on simulated time.
-
Algorithm State Setup (ISetupHandler)
Configure the algorithm cash, portfolio and data requested. Initialize all state parameters required.
These are all configurable from the config.json file in the Launcher Project.
Installation Instructions
We recommend using the docker image. This is perfectly configured to run out of the box without interfering with your development environment. You can pull this image with docker pull quantconnect/lean
Refer to the following readme files for a detailed guide regarding using our docker image with your local IDE:
To install locally, download the zip file with the latest master and unzip it to your favorite location. Alternatively, install Git and clone the repo:
git clone https://github.com/QuantConnect/Lean.git
cd Lean
macOS
- Install Visual Studio for Mac
- Open
QuantConnect.Lean.slnin Visual Studio
Visual Studio will automatically start to restore the Nuget packages. If not, in the menu bar, click Project > Restore NuGet Packages.
- In the menu bar, click
Run > Start Debugging.
Alternatively, run the compiled exe file. First, in the menu bar, click Build > Build All, then:
cd Lean/Launcher/bin/Debug
mono QuantConnect.Lean.Launcher.exe
Linux (Debian, Ubuntu)
- Install Mono:
sudo apt-get update && sudo rm -rf /var/lib/apt/lists/*
sudo apt-key adv --keyserver hkp://keyserver.ubuntu.com:80 --recv-keys 3FA7E0328081BFF6A14DA29AA6A19B38D3D831EF
echo "deb http://download.mono-project.com/repo/ubuntu stable-xenial/snapshots/5.12.0.226 main" > /etc/apt/sources.list.d/mono-xamarin.list && \
apt-get update && apt-get install -y binutils mono-complete ca-certificates-mono mono-vbnc nuget referenceassemblies-pcl && \
apt-get install -y fsharp && rm -rf /var/lib/apt/lists/* /tmp/*
If you get this error on the last command:
Unable to locate package referenceassemblies-pcl,
run the following command (it works on current version of Ubuntu - 17.10):
echo "deb http://download.mono-project.com/repo/ubuntu xenial main" | sudo tee /etc/apt/sources.list.d/mono-official.list
sudo apt-get update
sudo apt-get install -y binutils mono-complete ca-certificates-mono referenceassemblies-pcl fsharp
- Install Nuget
sudo apt-get update && sudo apt-get install -y nuget
- Restore NuGet packages then compile:
nuget restore QuantConnect.Lean.sln
msbuild QuantConnect.Lean.sln
If you get: "Error initializing task Fsc: Not registered task Fsc." -> sudo apt-get upgrade mono-complete
If you get: "XX not found" -> Make sure Nuget ran successfully, and re-run if neccessary.
If you get: "Confirm ... '.../QuantConnect.XX.csproj.*.props' is correct, and that the file exists on disk." -> Ensure that your installation path is free of reserved characters
If you get other errors that lead to the failure of your building, please refer to the commands in "DockerfileLeanFoundation" file for help.
- Run the compiled
exefile:
cd Launcher/bin/Debug
mono ./QuantConnect.Lean.Launcher.exe
- Interactive Brokers set up details
Make sure you fix the ib-tws-dir and ib-controller-dir fields in the config.json file with the actual paths to the TWS and the IBController folders respectively.
If after all you still receive connection refuse error, try changing the ib-port field in the config.json file from 4002 to 4001 to match the settings in your IBGateway/TWS.
Windows
- Install Visual Studio
- Open
QuantConnect.Lean.slnin Visual Studio - Build the solution by clicking Build Menu -> Build Solution (this should trigger the Nuget package restore)
- Press
F5to run
Nuget packages not being restored is the most common build issue. By default Visual Studio includes NuGet, if your installation of Visual Studio (or your IDE) cannot find DLL references, install Nuget, run nuget on the solution and re-build the Solution again.
Python Support
A full explanation of the Python installation process can be found in the Algorithm.Python project.
Local-Cloud Hybrid Development.
You can develop in your IDE and synchronize to the cloud with Skylight. For more information please see the Skylight Home.
Issues and Feature Requests
Please submit bugs and feature requests as an issue to the Lean Repository. Before submitting an issue please read others to ensure it is not a duplicate.
Mailing List
The mailing list for the project can be found on LEAN Forum. Please use this to request assistance with your installations and setup questions.
Contributors and Pull Requests
Contributions are warmly very welcomed but we ask you to read the existing code to see how it is formatted, commented and ensure contributions match the existing style. All code submissions must include accompanying tests. Please see the contributor guide lines.
All accepted pull requests will get a 2mo free Prime subscription on QuantConnect. Once your pull-request has been merged write to us at support@quantconnect.com with a link to your PR to claim your free live trading. QC <3 Open Source.
Acknowledgements
The open-sourcing of QuantConnect would not have been possible without the support of the Pioneers. The Pioneers formed the core 100 early adopters of QuantConnect who subscribed and allowed us to launch the project into open source.
Ryan H, Pravin B, Jimmie B, Nick C, Sam C, Mattias S, Michael H, Mark M, Madhan, Paul R, Nik M, Scott Y, BinaryExecutor.com, Tadas T, Matt B, Binumon P, Zyron, Mike O, TC, Luigi, Lester Z, Andreas H, Eugene K, Hugo P, Robert N, Christofer O, Ramesh L, Nicholas S, Jonathan E, Marc R, Raghav N, Marcus, Hakan D, Sergey M, Peter McE, Jim M, INTJCapital.com, Richard E, Dominik, John L, H. Orlandella, Stephen L, Risto K, E.Subasi, Peter W, Hui Z, Ross F, Archibald112, MooMooForex.com, Jae S, Eric S, Marco D, Jerome B, James B. Crocker, David Lypka, Edward T, Charlie Guse, Thomas D, Jordan I, Mark S, Bengt K, Marc D, Al C, Jan W, Ero C, Eranmn, Mitchell S, Helmuth V, Michael M, Jeremy P, PVS78, Ross D, Sergey K, John Grover, Fahiz Y, George L.Z., Craig E, Sean S, Brad G, Dennis H, Camila C, Egor U, David T, Cameron W, Napoleon Hernandez, Keeshen A, Daniel E, Daniel H, M.Patterson, Asen K, Virgil J, Balazs Trader, Stan L, Con L, Will D, Scott K, Barry K, Pawel D, S Ray, Richard C, Peter L, Thomas L., Wang H, Oliver Lee, Christian L.