[Rivet-announce] Rivet 3.0.0alpha1 released

[Chris Pollard]

Hello Rivet users,

We're happy to announce the availability of the first alpha of the next
major Rivet release, version 3. To grab it for yourself, please see the
Rivet downloads page,

https://www.hepforge.org/downloads/rivet,

or run the corresponding bootstrap script in the usual way

$ wget
http://rivet.hepforge.org/hg/bootstrap/raw-file/3.0.0alpha1/rivet-bootstrap

The primary new feature of the alpha release is the automatic handling of
MC events that a) have more than one HepMC event weight and/or b) are
comprised of multiple weighted "subevents."

In the case of a): if more than one weight is available in the HepMC file
(from different PDF sets, for example), then Rivet will create a histogram
for each of these weights, fill them appropriately, and write them out to
the relevant YODA file. Plotting scripts have been updated to show each of
these simultaneously, with the "nominal" histogram in bold.

This involves some non-trivial bookkeeping, and currently *only Histo1Ds
support multiple event weights*. The other AnalysisObjects will follow
soon, but we want to make this new feature available to our users so that
we can receive feedback before the final release.

A few *important* notes about the alpha:

- Analyses added after Rivet 2.2 are not included in the alpha but will, of
course, be available in the Rivet 3 release.
- A few analyses have been disabled because of how they handle the event
weight in the analyze() method; these analyses will be included in the 3.0
release.
- The alpha requires the latest version of YODA: 1.5.8.

The following additional points are important for those that are developing
new analyses:

- You should no longer explicitly include the event weight in
Histo1D::fill() calls; any weight passed to fill() will be multiplied by
the HepMC event weight when filling the histogram!
- Any Histo1D that is accessed in the Analysis::analyze() and
Analysis::finalize() methods should now be created using bookHisto1D in the
Analysis::init() method.

Please enjoy and do give feedback if you have comments, criticisms, or
suggestions.

Cheers,

Chris, for the Rivet developers
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://www.hepforge.org/lists-archive/rivet-announce/attachments/20160127/411ae178/attachment.html>