Feature #284

=== Fit ===

Added by wuttke over 6 years ago. Updated 10 months ago.

Status:RejectedStart date:09 Oct 2012
Priority:HighDue date:
Assignee:-% Done:

90%

Category:-
Target version:-

Description

This is the parent for all issues concerned with fitting.

Before investing substantial effort into improved fitting, we need to decide whether fitting will be done at C++ or Python level.


Subtasks

Bug #249: FunctionalTests: fitting from python works two times faster than fitting from C++Resolvedpospelov

Feature #151: Check with ToyExperiment the influence of numerical error in chi2/residuals on fit stabilityRejected

Feature #119: Fitting: implement boostrapping fit strategy to deal with local minima during the fitRejected

Feature #137: Fitting: implement strategy utilizing Gelfand algorithm to get out of local minimaRejected

Feature #101: Move ROOT minimization into ThirdParty and so get rid from ROOT dependencyResolvedpospelov

Feature #446: Repair broken Genetic minimizer after we get rid from ROOT dependencyArchivedpospelov

Feature #586: Investigate chi2-like objective functionsRejected

Feature #587: Investigate objective function alternatives.Long Term Idea

Feature #720: Provide OpenMPI support in FitSuiteRejected

Feature #767: Add functionality into MinimizerFactory::printCatalogue.Rejected

Feature #1055: Implement Genetic minimizer to get rid from ROOT dependencyArchivedpospelov

Feature #1056: Make default build to not to use system's ROOTArchivedherck

Bug #1210: GSLLMA minimizer sometimes doesn't respect fit parameter limits attributeRejected

Refactoring #1466: Refactor core Minimizer family to match the GUI presentationArchivedpospelov

Refactoring #1469: Package Multifit-Suite as separate libraryRejected

Refactoring #1542: Make ParameterPool and RealParameter independent of IParameterizedArchivedwuttke

Refactoring #1548: choose name for fit libraryRejected

Feature #1543: decouple FitKernel from FitSuiteArchivedwuttke

Refactoring #1547: restore ScanningMinimizerBacklog

Refactoring #1551: rename directory RootWrapper/ and files thereinResolved

Refactoring #1555: invert dependence of FitKernel on FitParameterLinkedResolvedwuttke

Refactoring #1557: FitParameterLinked causes weird couplingRejected

Refactoring #1588: FitSuiteParameters inhibits Python iteratorArchivedpospelov

Feature #1624: Make all reasonable parameter values into fittable parameter Resolvedpospelov

Refactoring #1650: Designation "ROOT" (or "Root" or "BA_ROOT") should not be used for code other than CERN ROOTResolved

Feature #1699: functional tests: Improve MinimizerCatalogue::toStringRejected

Refactoring #1700: Implement FitSuite::printParameters (or ::parametersToString.)Rejected

Feature #1702: Fit: Provide extended print output for fit parameters during iterations.Resolvedpospelov

Feature #1704: Revise chi squared moduleNew

Bug #1705: Py Fit crashes for zero fit parametersRejected

History

#1 Updated by wuttke over 3 years ago

  • Tracker changed from Feature to Envelope task
  • Status changed from Backlog to In Progress

#2 Updated by wuttke over 3 years ago

  • Subject changed from Fit campaign to *Fit campaign*

#3 Updated by wuttke over 3 years ago

  • Subject changed from *Fit campaign* to == Fit campaign ==

#4 Updated by wuttke over 3 years ago

  • Subject changed from == Fit campaign == to === Fit ===

#5 Updated by pospelov about 2 years ago

  • Tracker changed from Envelope task to Feature
  • Status changed from In Progress to Rejected

Also available in: Atom PDF