The main point for this branch is to do some core change related to the
EFT validation effort (lead by Gauthier/Fabio). This is implemented as a plugin but the accord required some change in the re-weighting module (mainly for the formatting).
Those changes adds:
1) The following command: import process PATH is now supported. (allowing to provide a file with all your process definition whithout any generate/add process
2) Allow the re-weighting to distinguish process based on the process event written in the lhe file / @X [does not seems to work for loop]
3) More xml like formatting for the weight definition (including model version/...)
4) the support for NHEL=-2 in reweighting, which returns the sum over polarization (like -1) but clear all filter associated to the zero-helicity to avoid any possible bias)
5) a (not documented) flag --no_warning=nodiagram which prevent the code to crash if you do not have any diagram generated for a given process
6) possibility to call the reweighting interface as a standalone executable
7) improve the templating of the standalone output to support model with more than 1k parameters (and reduce RAM use for model with less parameters)
The main point for this branch is to do some core change related to the
EFT validation effort (lead by Gauthier/Fabio). This is implemented as a plugin but the accord required some change in the re-weighting module (mainly for the formatting).
Those changes adds: nodiagram which prevent the code to crash if you do not have any diagram generated for a given process
1) The following command: import process PATH is now supported. (allowing to provide a file with all your process definition whithout any generate/add process
2) Allow the re-weighting to distinguish process based on the process event written in the lhe file / @X [does not seems to work for loop]
3) More xml like formatting for the weight definition (including model version/...)
4) the support for NHEL=-2 in reweighting, which returns the sum over polarization (like -1) but clear all filter associated to the zero-helicity to avoid any possible bias)
5) a (not documented) flag --no_warning=
6) possibility to call the reweighting interface as a standalone executable
7) improve the templating of the standalone output to support model with more than 1k parameters (and reduce RAM use for model with less parameters)