what is "WARNING: Failed to update dependent parameter"

Created by Olivier Mattelaer on on 2016-12-21
Keywords:
Last updated by:
Olivier Mattelaer on on 2016-12-21

In the param_card some parameters are not free parameters.
Some of them have a fixed value (like the mass of massless particle)
Some of them have a value defined by an analytical formula (like the W mass in our default EW scheme)

We never use the input of those parameter for the computation of the matrix-element/cross-section/generation of events.
But some code like the parton-shower or the decay package (MadSpin) can use the value written in the param_card even if those value are inconsistent.

We therefore try to automatically update such parameters before the computation of the cross-section. Such that such parameter will be consistently defined when used by sub-sequent program.

However
1) we do not want to significantly slow down the code by performing such operation (and the loading of the model can be very long)
2) The loading of the model has some border effect which can prevent to update to run if a previous update was performed on another model
In both case, (too slow/border effect detected) the update will raise the warning:
"""WARNING: Failed to update dependent parameter. This might create trouble for external program (like MadSpin/shower/...)"""
In that case if you use any program after the event generation, you need to be sure that all your dependent parameter are consistently define (or that you do not use those).

If your model takes too long to load and that you still want to use the functionality, you can type (at the time of the question for the handling of the card) "update dependent"