[Rivet-announce] Professor 2.2.1

[Holger Schulz]

Dear tuning and parametrisation enthusiasts,

we are very pleased to announce the release of Professor 2.2.1

This release focuses on robustness when reading in files for parametrisation.

In previous releases we chose exit the parametrisation stage on encountering
nan coefficients which are a symptom of very low stats in the input histograms.

We now print out a warning and simply ignore the histograms in question.
This approach is much more pragmatic especially in cases of many hundred
histograms per parameter point.

You can (and should) always check the observables available in a parametrisation
output file using prof2-ls.

We changed the default backend of matplotlib to Agg in the residuals, envelopes and
sensitivity scripts to allow producing plots in the absence of X11.

This is of great use when working with docker remotely.

And of course, we provide a docker image ready to be pulled via 
docker pull iamholger/professor:2.2.1

The image comes with yoda 1.6.6, root support and is based on fedora25.

We added some documentation to the wiki that shows how you can download and
get started with professor in 3 simple steps:

    https://professor.hepforge.org/trac/wiki/docker

Thanks,
Holger and Andy
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet-announce/attachments/20170203/1f6414f9/attachment.html>