Files
quantconnect--lean/Tests/Python/PythonWrapperTests.cs
T
Martin-Molinero 27de93f78f
Syntax Tests / build (push) Has been cancelled
API Tests / build (push) Has been cancelled
Benchmarks / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
Regression Tests / build (push) Has been cancelled
Report Generator Tests / build (push) Has been cancelled
Research Regression Tests / build (push) Has been cancelled
Python Virtual Environments / build (push) Has been cancelled
Market orders wait for fresh data instead of filling on stale prices (#9535)
* Market orders wait for fresh data instead of filling on stale prices

A market order would previously fill immediately on the most recent
available data even when that data was older than StalePriceTimeSpan
(default one hour), only attaching a warning. This is unrealistic for a
coarse resolution asset (hour/daily) where the latest bar is the stale
previous close when the order is placed mid-bar or via an intraday
scheduled event.

The default fill models (FillModel, EquityFillModel, FutureFillModel) now
wait for fresh data instead of filling on a stale price, but only for hour
and daily resolutions; the order fills when the next bar closes. For
minute/second/tick subscriptions the previous behavior is kept (fill on
the stale price with a warning), since stale data there is a genuine gap
rather than a bar still forming.

Adds HourResolutionMarketOrderStalePriceRegressionAlgorithm, updates the
FillOutsideHours daily expectation, and regenerates statistics for the
hour/daily algorithms whose fills change. FutureOptionDaily buys and
liquidates a day apart now (a same-day buy + liquidate cannot fill on
daily data once stale fills are disabled).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Normalize and correct StalePriceTimeSpan XML docs

The interface and class docs now match and reflect the actual behavior:
the wait-for-fresh-data only applies to hour/daily resolutions, while
minute/second/tick subscriptions still fill on stale data with a warning.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Fill resting market orders at the bar open instead of the close

A hour/daily market order that was resting before the current bar opened
(it predates the bar - placed after the previous close or while waiting
for fresh data) now fills at the bar open, the price when trading resumed
(like a MarketOnOpen), instead of the bar close. Orders placed during the
bar still fill at the current/close price, so intraday mid-bar fills are
unchanged. Equity fills are unchanged (resting equity orders are already
converted to MarketOnOpen by QCAlgorithm.MarketOrder).

Adds the shared FillModel.GetMarketFillPrice helper used by the base
FillModel and FutureFillModel, a unit test, and regenerates statistics for
the affected daily/hour futures, index and crypto regression algorithms
(order counts unchanged, only fill prices).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Add regression algorithm asserting resting market orders fill at the bar open

RestingMarketOrderFillsAtBarOpenRegressionAlgorithm buys a daily future on the
bar that delivers it (fills at that bar's close) and submits a liquidation while
the market is closed (overnight pulse, no fresh bar). The liquidation rests and
fills on a later bar at the bar open, not its close - asserting the new
GetMarketFillPrice behavior. The in-bar buy is asserted to fill at the close, for
contrast.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Carry the bar start time on Prices instead of re-reading the cache

Add Prices.Time (the bar start, mirroring BaseData.Time/EndTime), populated from
the source bar/tick in every GetPrices path. GetMarketFillPrice now uses
prices.Time directly instead of a second asset.Cache.GetData() lookup. Behavior
is unchanged (prices.Time equals the previously read cache time).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Add regression algorithm asserting in-session hour orders fill at the latest close

HourMarketOrderFillsAtBarCloseRegressionAlgorithm submits an hour resolution
market order mid-bar (via an intraday scheduled event) while the market is open,
using the default one hour StalePriceTimeSpan. It asserts the order fills
immediately at the latest available bar's close - not waiting and not at the bar
open - since the latest bar is within the stale window. Guards the resting-order
open-fill behavior against affecting ordinary in-session fills.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Regenerate custom fill model algorithm statistics for the open-fill change

CustomModelsAlgorithm and CustomPartialFillModelAlgorithm subscribe SPY at hour
resolution and their custom fill models delegate to base.MarketFill, so resting
orders now fill at the bar open. Regenerate their statistics (C#/Python) and the
inline expected statistics of the PEP8StyleCustomModelsWork test.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
2026-06-18 11:59:19 -03:00

392 lines
14 KiB
C#

/*
* QUANTCONNECT.COM - Democratizing Finance, Empowering Individuals.
* Lean Algorithmic Trading Engine v2.0. Copyright 2014 QuantConnect Corporation.
*
* Licensed under the Apache License, Version 2.0 (the "License");
* you may not use this file except in compliance with the License.
* You may obtain a copy of the License at http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
using System;
using NUnit.Framework;
using Python.Runtime;
using QuantConnect.Python;
using System.Collections.Generic;
using QuantConnect.Statistics;
using System.Reflection;
namespace QuantConnect.Tests.Python
{
public static class PythonWrapperTests
{
[TestFixture]
public class ValidateImplementationOf
{
[TestCase(nameof(MissingMethodOne), "ModelMissingMethodOne", "MethodOne")]
[TestCase(nameof(MissingProperty), "ModelMissingProperty", "PropertyOne")]
public void ThrowsOnMissingMember(string moduleName, string className, string missingMemberName)
{
using (Py.GIL())
{
var moduleStr = GetFieldValue(moduleName);
var module = PyModule.FromString(nameof(ValidateImplementationOf), moduleStr);
var model = module.GetAttr(className).Invoke();
Assert.That(() => model.ValidateImplementationOf<IModel>(), Throws
.Exception.InstanceOf<NotImplementedException>().With.Message.Contains(missingMemberName));
}
}
[TestCase(nameof(FullyImplemented), "FullyImplementedModel")]
[TestCase(nameof(FullyImplementedSnakeCase), "FullyImplementedSnakeCaseModel")]
[TestCase(nameof(FullyImplementedWithPropertyAsField), "FullyImplementedModelWithPropertyAsField")]
[TestCase(nameof(DerivedFromCsharp), "DerivedFromCSharpModel")]
public void DoesNotThrowWhenInterfaceFullyImplemented(string moduleName, string className)
{
using (Py.GIL())
{
var moduleStr = GetFieldValue(moduleName);
var module = PyModule.FromString(nameof(ValidateImplementationOf), moduleStr);
var model = module.GetAttr(className).Invoke();
Assert.That(() => model.ValidateImplementationOf<IModel>(), Throws.Nothing);
}
}
[Test]
public void SettlementModelPythonWrapperWorks()
{
var results = AlgorithmRunner.RunLocalBacktest("CustomSettlementModelRegressionAlgorithm",
new Dictionary<string, string>()
{
{PerformanceMetrics.TotalOrders, "0"},
{"Average Win", "0%"},
{"Average Loss", "0%"},
{"Compounding Annual Return", "119.460%"},
{"Drawdown", "0%"},
{"Expectancy", "0"},
{"Net Profit", "1.010%"},
{"Sharpe Ratio", "-5.989"},
{"Sortino Ratio", "0"},
{"Probabilistic Sharpe Ratio", "1.216%"},
{"Loss Rate", "0%"},
{"Win Rate", "0%"},
{"Profit-Loss Ratio", "0"},
{"Alpha", "-0.411"},
{"Beta", "-0.033"},
{"Annual Standard Deviation", "0.079"},
{"Annual Variance", "0.006"},
{"Information Ratio", "-10.086"},
{"Tracking Error", "0.243"},
{"Treynor Ratio", "14.619"},
{"Total Fees", "$0.00"},
{"Estimated Strategy Capacity", "$0"},
{"Lowest Capacity Asset", ""},
{"Portfolio Turnover", "0%"},
{"OrderListHash", "d41d8cd98f00b204e9800998ecf8427e"}
},
Language.Python,
AlgorithmStatus.Completed,
algorithmLocation: "../../../Algorithm.Python/CustomSettlementModelRegressionAlgorithm.py"
);
}
[Test]
public void BenchmarkModelPythonWrapperWorks()
{
var results = AlgorithmRunner.RunLocalBacktest("CustomBenchmarkRegressionAlgorithm",
new Dictionary<string, string>()
{
{PerformanceMetrics.TotalOrders, "0"},
{"Average Win", "0%"},
{"Average Loss", "0%"},
{"Compounding Annual Return", "0%"},
{"Drawdown", "0%"},
{"Expectancy", "0"},
{"Net Profit", "0%"},
{"Sharpe Ratio", "0"},
{"Sortino Ratio", "0"},
{"Probabilistic Sharpe Ratio", "0%"},
{"Loss Rate", "0%"},
{"Win Rate", "0%"},
{"Profit-Loss Ratio", "0"},
{"Alpha", "0"},
{"Beta", "0"},
{"Annual Standard Deviation", "0"},
{"Annual Variance", "0"},
{"Information Ratio", "-6.654540153820717E+27"},
{"Tracking Error", "11.906"},
{"Treynor Ratio", "0"},
{"Total Fees", "$0.00"},
{"Estimated Strategy Capacity", "$0"},
{"Lowest Capacity Asset", ""},
{"Portfolio Turnover", "0%"},
{"OrderListHash", "d41d8cd98f00b204e9800998ecf8427e"}
},
Language.Python,
AlgorithmStatus.Completed,
algorithmLocation: "../../../Algorithm.Python/CustomBenchmarkRegressionAlgorithm.py"
);
}
[Test]
public void PEP8StyleAlgorithmsImplementationsWork()
{
AlgorithmRunner.RunLocalBacktest("PEP8StyleBasicAlgorithm",
new Dictionary<string, string>()
{
{"Total Orders", "1"},
{"Average Win", "0%"},
{"Average Loss", "0%"},
{"Compounding Annual Return", "271.453%"},
{"Drawdown", "2.200%"},
{"Expectancy", "0"},
{"Start Equity", "100000"},
{"End Equity", "101691.92"},
{"Net Profit", "1.692%"},
{"Sharpe Ratio", "8.854"},
{"Sortino Ratio", "0"},
{"Probabilistic Sharpe Ratio", "67.609%"},
{"Loss Rate", "0%"},
{"Win Rate", "0%"},
{"Profit-Loss Ratio", "0"},
{"Alpha", "-0.005"},
{"Beta", "0.996"},
{"Annual Standard Deviation", "0.222"},
{"Annual Variance", "0.049"},
{"Information Ratio", "-14.565"},
{"Tracking Error", "0.001"},
{"Treynor Ratio", "1.97"},
{"Total Fees", "$3.44"},
{"Estimated Strategy Capacity", "$56000000.00"},
{"Lowest Capacity Asset", "SPY R735QTJ8XC9X"},
{"Portfolio Turnover", "19.93%"},
{"OrderListHash", "3da9fa60bf95b9ed148b95e02e0cfc9e"}
},
Language.Python,
AlgorithmStatus.Completed,
algorithmLocation: "../../../Algorithm.Python/PEP8StyleBasicAlgorithm.py"
);
}
[Test]
public void PEP8StyleCustomModelsWork()
{
AlgorithmRunner.RunLocalBacktest("CustomModelsPEP8Algorithm",
new Dictionary<string, string>()
{
{"Total Orders", "63"},
{"Average Win", "0.10%"},
{"Average Loss", "-0.06%"},
{"Compounding Annual Return", "-7.101%"},
{"Drawdown", "2.400%"},
{"Expectancy", "-0.181"},
{"Start Equity", "100000"},
{"End Equity", "99383.07"},
{"Net Profit", "-0.617%"},
{"Sharpe Ratio", "-1.441"},
{"Sortino Ratio", "-1.977"},
{"Probabilistic Sharpe Ratio", "22.128%"},
{"Loss Rate", "69%"},
{"Win Rate", "31%"},
{"Profit-Loss Ratio", "1.64"},
{"Alpha", "-0.101"},
{"Beta", "0.121"},
{"Annual Standard Deviation", "0.04"},
{"Annual Variance", "0.002"},
{"Information Ratio", "-4.109"},
{"Tracking Error", "0.102"},
{"Treynor Ratio", "-0.475"},
{"Total Fees", "$62.23"},
{"Estimated Strategy Capacity", "$52000000.00"},
{"Lowest Capacity Asset", "SPY R735QTJ8XC9X"},
{"Portfolio Turnover", "197.93%"},
{"OrderListHash", "fe01fe4923e8856fe3376ece636b4e23"}
},
Language.Python,
AlgorithmStatus.Completed,
algorithmLocation: "../../../Algorithm.Python/CustomModelsPEP8Algorithm.py"
);
}
private static string GetFieldValue(string name)
{
return typeof(ValidateImplementationOf).GetField(name, BindingFlags.Static | BindingFlags.NonPublic).GetValue(null) as string;
}
private const string FullyImplemented =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class FullyImplementedModel:
def MethodOne():
pass
def MethodTwo():
pass
@property
def PropertyOne(self):
return 'value'
";
private const string FullyImplementedSnakeCase =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class FullyImplementedSnakeCaseModel:
def method_one():
pass
def method_two():
pass
@property
def property_one(self):
pass
";
private const string FullyImplementedWithPropertyAsField =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class FullyImplementedModelWithPropertyAsField:
def method_one():
pass
def method_two():
pass
def __init__(self):
self.property_one = 'value'
";
private const string DerivedFromCsharp =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class DerivedFromCSharpModel(PythonWrapperTests.ValidateImplementationOf.Model):
def MethodOne():
pass
@property
def PropertyOne(self):
return 'value'
";
private const string MissingMethodOne =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class ModelMissingMethodOne:
def MethodTwo():
pass
@property
def PropertyOne(self):
return 'value'
";
private const string MissingProperty =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class ModelMissingProperty:
def MethodOne():
pass
def MethodTwo():
pass
";
interface IModel
{
string PropertyOne { get; set; }
void MethodOne();
void MethodTwo();
}
public class Model : IModel
{
public string PropertyOne { get; set; }
public void MethodOne()
{
}
public void MethodTwo()
{
}
}
}
[TestFixture]
public class InvokeTests
{
[Test]
public void InvokesCSharpMethod()
{
using (Py.GIL())
{
var module = PyModule.FromString(nameof(InvokeTests), InvokeModule);
var model = module.GetAttr("PythonInvokeTestsModel").Invoke();
Assert.That(model.InvokeMethod<int>("AddThreeNumbers", 1, 2, 3), Is.EqualTo(6));
}
}
[Test]
public void InvokesPythonMethod()
{
using (Py.GIL())
{
var module = PyModule.FromString(nameof(InvokeTests), InvokeModule);
var model = module.GetAttr("PythonInvokeTestsModel").Invoke();
Assert.That(model.InvokeMethod<int>("AddTwoNumbers", 1, 2), Is.EqualTo(3));
}
}
private const string InvokeModule =
@"
from clr import AddReference
AddReference('QuantConnect.Tests')
from QuantConnect.Tests.Python import *
class PythonInvokeTestsModel(PythonWrapperTests.InvokeTests.InvokeTestsModel):
def add_two_numbers(self, a, b):
return a + b
";
public class InvokeTestsModel
{
public int AddTwoNumbers(int a, int b)
{
throw new NotImplementedException();
}
public int AddThreeNumbers(int a, int b, int c)
{
return a + b + c;
}
}
}
}
}