Cargando…
Reproducing searches for new physics with the ATLAS experiment through publication of full statistical likelihoods
The ATLAS Collaboration is starting to publicly provide likelihoods associated with statistical fits used in searches for new physics on HEPData. These likelihoods adhere to a specification first defined by the $\texttt{HistFactory}$ p.d.f. template. This note introduces a JSON schema that fully des...
Autor principal: | |
---|---|
Lenguaje: | eng |
Publicado: |
2019
|
Materias: | |
Acceso en línea: | http://cds.cern.ch/record/2684863 |
Sumario: | The ATLAS Collaboration is starting to publicly provide likelihoods associated with statistical fits used in searches for new physics on HEPData. These likelihoods adhere to a specification first defined by the $\texttt{HistFactory}$ p.d.f. template. This note introduces a JSON schema that fully describes the $\texttt{HistFactory}$ statistical model and is sufficient to reproduce key results from published ATLAS analyses. This is per-se independent of its implementation in $\texttt{ROOT} $and it can be used to run statistical analysis outside of the $\texttt{ROOT}$ and $\texttt{RooStats}$/$\texttt{RooFit}$ framework. The first of these likelihoods published on HEPData is from a search for bottom-squark pair production. Using two independent implementations of the model, one in $\texttt{ROOT}$ and one in pure Python, the limits on the bottom-squark mass are reproduced, underscoring the implementation independence and long-term viability of the archived data. |
---|