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
API Tests / build (push) Has been cancelled
* Fix half of the CA1051 warnings
This warning is about not declaring visible instance fields. There are
something about 500 warnings in the solution, mostly in the QuantConnect and QuantConnect.Algorithm.CSharp projects. I aim to fix one of them in this PR and the other half of them in a second one. To fix it, I'm changing the visible instancce fields for properties.
* fix bugs
* Addressing minor reviews
* More minor fixes
---------
Co-authored-by: Martin Molinero <martin.molinero1@gmail.com>
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
* Plotting Limits Handling Improvements
- Improve the handling and enforcement of plotting limits
- Increase local default limits
* Minor plotting test fix
Regression Tests / build (push) Has been cancelled
Build & Test Lean / build (push) Has been cancelled
* Remove user plan enum
- Get maximum order and runtime from job controls
- Remove user plan enum definitions
* Add log for BacktestingSetupHandler maximums
* initial commit
* run parametrized algorithm with command line parameters
* skeleton: top level structure
* OptimizationNodePacket scheme
* pass parameters as HashSet
* run Lean and read results
* call method on optimization completion
* refactor public interfaces
- close ParameterSet collection; allow only get operations
- explicit method to start LeanOptimizer
* synchronize RunLean method; the result could come in before the backtest id is set in the collections
* another portion of refactoring and interface changes
* comments
* comments & tests for Extremum, Minimization and Maximization classes
* unify optimization paramater values (min, max, step) & mode GridSearch tests
- swap min&max if necessary
- iterate left => right (negate step value if necessary) & provide default step value if step == 0
- no StackOverflow Exception
- parameterSet Id should be global for current generator and retain between steps
- test signle point boundary (min == max)
* BruteForceStrategy tests
* more comments
* Update Optimizer assembly information
- Update Optimizer projects assembly information to match behavior of
the other projects
* Tweaks
- Adding comments
- Replace OnComplete for Ended event
- Replace Abort for Dispose
- ConsoleLeanOptimizer will keep track of running processes
- Each backtest will store results in a separated directory, so they
don't fight for the log.txt file.
- Adding cmdline option for lean to close automatically
- Adding concurrent execution backtest limit
- Console optimizer will start Lean minimized
- Escape spaces in Json path
* remove parameter set generator abstraction layer
we don't need this flexibility now.
* refactor public methods; Step shouldn't be public
* constraints: wip
* define contract
* comparison operators and tests
* specify JsonProperty values
* Move SafeMultiply100 to extensions
* Throw exception on failed Optimizer.Start
* constraints: wip
* change finish & dispose process
* minor fixes
- handle force lean abort
- notify consumer if target has been reached
* target & constraints; adapt unit tests
* Minor Tweaks and fixes
- Some logging improvements
- Remove Public since not required
* Ignore empty ParameterValue
* simplify condition
* avoid reinitialization
* reduce type; force immutable
* unit tests for constraints and target value
* parse & normalize percent values, i.e. 20% => 0.2
* fixup
* Target & Constraint & OptimizationNodePacket unit tests
* Add more json unit tests
- Adding more json conversion unit tests. Fix bug for Extremum which
wasn't using the converter.
* LeanOptimizer tests
* Estimation results
* User thread safe counters
* LeanOptimizer unit tests; push OptimizationResult on Ended event
* more unit tests
* Minor tweaks
-Estimate ToString in a single line.
-Typos and missing header file
* Add base SendUpdate method
- Add base SendUpdate method for LeanOptimizer
* fix LeanOptimizer test; rely on internal Update rather than timer
* Add OptimizationStatus
- Add missing commments and OptimizationStatus
* EulerSearch implementation: wip
* OptimizationParameter custom converter
* change the type
* make step optional
* change folder structure
* enumerate optimization parameter using IEnumerable & IEnumerator
* unit tests: parameters & objectives
* unit tests: strategies
* remove redundant TODO
* change Euler search boundaries
* more Euler tests
* prevent race condition
* Add account/read endpoint
- Adding account/read endpoint. Adding unit test
* Add status check before running lean
* Minor self review
- Adding missing comments, minor changes
* remove array parameters
* minor changes
- tidy up config file, rename variable
- accept min less or equal than max
* move OptimizationParameter methods to strategies
* Minor improvements for BaseResultHandler derivates
* minor changes
- strict requirements for Step and MinStep values
- strategy specific settigs
* Add TotalRuntime to estimate
Co-authored-by: Martin Molinero <martin.molinero1@gmail.com>
- Adding new DataPermissionManager that will own the existing
datachannelProvider. Will assert configurations before added to the
data feed. Adding unit tests
Errors raised during persistence aren't able to be handled by user code,
and in fact, are swallowed by the implementation after being logged. By
exposing these errors as events we allow the algorithm to be notified of
such an error and take any step necessary to handle the persistence error.
persistenceIntervalSeconds defines the number of seconds between
each save operation. For the local object store, this dictates
how often the contents of the object store is packaged and written
to disk. The PersistData virtual method is provided for subclasses
to provide a different implementation of how/where to persist the
data. The change to be in-memory aims at keeping the object store
performant with reasonable persistence guarantees.
The job packet is now able to specify a maximum file count as well
as a maximum storage size limit. These controls are enforced in the
LocalObjectStorage implementation.
We restrict each algorithm time loop to a pre-determined amount of time.
Exceeding this limit will cause the algorithm to immediately terminate.
This quickly becomes an issue when considering users running trainable
models that have a long initialization period that exceeds the time loop
maximum.
This change provides a mechanism through which a long-running scheduled
event is permitted to keep running and is permitted to avoid the time loop
permitted by requesting additional time. Requests for additional time are
limited according to a leaky bucket implementation whose parameters are
set via the job's controls structure. The fundamental time unit for the
algorithm is a single minute.
Here's how it works. If a scheduled event takes longer than one full wall
clock second then a request is made to the leaky bucket for one more minute.
If the scheduled event continues to take more time, it will continue to
request additional minutes. Each requested minute will prevent the algorithm's
time loop check from terminating the algorithm. When the bucket is empty and
no more minutes are available to be requested, a TimeoutException is thrown
causing a cascade that ends in the algorithm's termination and status being
flipped to RuntimeError.
Additionally, this applies equally to ALL scheduled events. While some helpers
were added with the naming of Train and TrainNow to the ScheduleManager, these
methods don't do anything special and the infrastructure doesn't otherwise
flag them as different, so this feature becomes part of the core Scheduled
Event feature set.
Further, the live scheduled events were not touched and are still pending
further discussion regarding the value added by enforcing a time restriction
when simulation time and wall clock time are equivalent.
Fixes#3319
The log limit will now be retrieved from the AlgorithmNodePacket when either a backtest or live algorithm uploads user logs at the end of the algorithm
Jhonathan Abreu