CalculateLimitToys

Perform a hypothesis test using the asymptotic-formulae approach (https://arxiv.org/abs/1007.1727), calculating the significiance of a discovery. This action does the same as CalculateLimit, but uses the CommonStatTools AsymptoticsCLsRunner as a backend. It is only available when CommonStatTools are available.

Usage:

+CalculateLimitToys {
    +HWWHighMass {
        <nPoints = 20, nToys = 50000, point = "1000">
    }
}

The options include:

• dataset: The name of the dataset to fit.
• point: name of the point (e.g. mass value)
• modelConfig (default=ModelConfig): The name of the ModelConfig to be used.
• calculatorType: (default=0) Type of the calculator to be used
  • 0: Freq calculator
  • 1: Hybrid calculator
  • 2: Asymptotic calculator
  • 3: Asymptotic calculator using nominal Asimov data sets (not using fitted parameter values but nominal ones)
• testStatType: (default=2). Type of the test statistic to be used
  • 0: LEP
  • 1: Tevatron
  • 2: Profile Likelihood two sided
  • 3: Profile Likelihood one sided (i.e. = 0 if mu < muhat)
  • 4: Profile Likelihood signed ( pll = -pll if mu < muhat)
  • 5: Max Likelihood Estimate as test statistic
  • 6: Number of observed event as test statistic
• poi.min: minimum value of the POI to be used
• poi.max: minimum value of the POI to be used
• nPoints: number of points to evalute the POI for
• useCLs: (default=true) toggle whether to use CLs
• confidenceLevel: (default=0.95) The confidence level for which to calculate the limit
• optimize (default=true): optimize evaluation of test statistic
• useVectorStore (default=true): convert data to use new roofit data store
• generateBinned (default=true):generate binned data sets
• nToysRatio (default=2): ratio Ntoys S+b/ntoysB
• poi.min (default=-1): minimum of POI to be used
• poi.max (default=-1): maximum of POI to be used
• useProof (default=false): use Proof to parallelize
• useProofWorkers (default=0): number of Proof workers
• detailed (default=false): enable detailed output
• reuseAltToys (default=false): reuse same toys for alternate hypothesis (if set one gets more stable bands)
• rebuildToys (default=false): re-do extra toys for computing expected limits and rebuild test stat distributions (N.B this requires much more CPU (factor is equivalent to nToyToRebuild)
• nToysRebuild (default=100): number of toys to rebuild
• rebuildParamValues (default=0):
• minimizerType (default=""): minimizer type (default is what is in ROOT::Math::MinimizerOptions::DefaultMinimizerType()
• initialFit (default=-1): do a first fit to the model (-1 : default, 0 skip fit, 1 do always fit)
• asimovBins (default=0): number of bins in observables used for Asimov data sets (0 is the default and it is given by workspace, typically is 100)