Use status 20 and status 3 partons in filters
Hi all, I'm working on a Sherpa filter in 2.2.2 that computes the dijet mass from partons. This is similar to an issue that was recently resolved in #86 (closed). One of the things I've noticed is that there are very long tails from low mjj (parton) to high mjj (truth), where truth is defined using jets after the parton shower. I think I have identified one of the reasons for this. It appears as though events that have status 20 partons ignore the status 3 partons. Note that my understanding of the Sherpa filter is the following: if you have a status 20 parton, only take the status 20 partons to compute the filter variable, else use status 3. The problem is, you only ever have two status 20 partons, and thus can miss a significant portion of the event.
You can see this in the attached event display. This is an event where if you take just the status 20 partons, you get MJJ = 150 GeV, but if you look at the jets in the event after the parton shower, mjj = 2.1 TeV. At first I thought that this was very odd looking, until I realized that one of the jets corresponds to a status 3 parton. Quantitatively, this 'tail' of high mjj events is mainly made up of this category. If you change the filter to do a simple overlap removal, e.g., take the status 20 partons and all status 3 partons that do not overlap with a status 3, then the tails virtually disappear.
Is there a way to modify the filter to include all of the partons in the filter, as in my naive emulation mentioned above?
Note that in the event display, the red circles are status 20 partons, blue are status 3 and black are truth jets. The partons are slightly shifted from one another in eta,phi to be visible...
Ben