This PR adds support for specifying the noise level (L0/L1) added to the projections directly from the command line.
Creating projections
Fits with projections follow a two-step process. First, one creates projected datasets (typically with reduced statistical
and systematic uncertainties) as .yaml files in the standard SMEFiT format with the following command,
where <luminosity> specifies the luminosity of the projection in :math:{\rm fb}^{-1}. The noise level <luminosity>
can be either L0 are L1 corresponding to either level 0 or level 1 projections respectively. In level 0 projections,
the experimental central value coincides exactly with the theory prediction, while the experimental central values are fluctuated around the theory prediction according to the experimental uncertainties in case of level 1. If <noise level> is not specified, level 0 is assumed. If <luminosity> is not specified, the original luminosities are kept and the uncertainties are not rescaled. The projection_runcard specifies which datasets need to be extrapolated, by which factor to reduce the systematics, and sets the necessary paths:
This PR adds support for specifying the noise level (L0/L1) added to the projections directly from the command line.
Creating projections
Fits with projections follow a two-step process. First, one creates projected datasets (typically with reduced statistical and systematic uncertainties) as
.yaml
files in the standard SMEFiT format with the following command,where
<luminosity>
specifies the luminosity of the projection in :math:{\rm fb}^{-1}
. The noise level<luminosity>
can be eitherL0
areL1
corresponding to either level 0 or level 1 projections respectively. In level 0 projections, the experimental central value coincides exactly with the theory prediction, while the experimental central values are fluctuated around the theory prediction according to the experimental uncertainties in case of level 1. If<noise level>
is not specified, level 0 is assumed. If<luminosity>
is not specified, the original luminosities are kept and the uncertainties are not rescaled. Theprojection_runcard
specifies which datasets need to be extrapolated, by which factor to reduce the systematics, and sets the necessary paths: