Event Display Scanning For Beginners¶
This is the wiki page for the NOvA Event Display Scanning. The intention is to provide answers to the questions that you might have as a newbie in the scanning effort. The goal is to catch event displays that have the potential to become a neutrino candidates after further analysis.
Starting (getting things ready)¶
- Instructions for Running the Scanning Display GUI:
- Assuming you are in one of the NOvA virtual machines (i.e. nova02), set up your environment using development release (i.e. setup_nova).
- Run the scanner by typing numi_scan and wait for the GUI to load. That should take a couple of seconds BUT less that a minute.
- After that, on the terminal you'll read the following:
Hello <username> I've found a file for you! * Please wait a moment while the event display starts. (You won't see the normal start up scrawl...) * When it starts, open Window->Scan Window. * Please place the scan window over event number in upper right corner. * Please scan all the way to the end of the file. * Please DO NOT use the Back button.
- In the upper left menu click on Window -> Scan Window and place it over the event number in the upper right corner (why?).
- When looking to an event, you can make decisions on it based in a combination of intuition, physics knowledge and perception. According to that, fill the charts in the Scan Window and/or write comments. You can also take a look at [Scanning Rules] for references.
- Once your decision on the current GUI event is made, DO NOT navigate using Previous/Next or forwards/back buttons, instead press Record in the Scan Window if you want to go to the next event.
- If you did a mistake and want to go back one event? (EXPECTED).
- You can always zoom in to a different XZ and YZ range by clicking and dragging on the number labels of the GUI. You can unzoom by giving left-click -> UnZoom or click on Reload button in the upper left (remember that to do so your pointer has to change into a 'hand').
- The aspect ratio of the events is not protected, so try to unzoom one last time before you click on Record to see the whole picture.
- Other options are available in the Edit menu on the upper left.
- To exit the GUI go to the upper left menu and click File->Quit.
NOTE: If for some reason when scanning your results become invalid (due to a network interruption, power outage, crash of the event display GUI, etc) try to send an e-mail to Gavin Davies (gsdavies@fnal.gov) or Michale Baird (mbaird42@fnal.gov) with your username and the time of your scan so that your results can be removed without any other complication.
Further reference to the Event Display GUI: Running the EventDisplay
EXAMPLE: 2013-12-06 Tutorial at FNAL¶
For the purpose of this tutorial and as an illustration, a sample file created by Michael Baird was used. It contains events that are not a sample for the actual neutrino scanning search so they should not be considered seriously.
Recalling the previous instructions: On a NOvA machine and setting development without actually doing 'srt_setup -a' you can open the sample file:
nova -c evd_numi_2db.fcl /nova/ana/users/mbaird42/numi_scan_example.root
After just a couple of seconds BUT less than one minute, you'll have the evd GUI active:
From this image we can appreciate that:
- The timing window in the bottom left has been removed.
- Blue solid line is a slice or cluster (bag) of hits, associated/correlated between themselves. The intention is that each slice corresponds to one neutrino interaction.
- Blue dotted lines represent tracks (collection of hits) with a reconstructed path through them, i.e. CosmicTrack in this case.
After is decided what is the possible nature of this event and once you have filled the Scan Window charts and/or writing some comments, the next step is click on Record for continue scanning the next event.
Scanning Rules (what to look and how)¶
- MC and Data: avoid becoming bias!
FarDet Event classification (tentatively).¶
- Typical event
- Logic Flow chart
- Technical Details
- Event Categories:
- Golden-the ones that are candidates for going to the press.
- Interesting-the ones that have something that make them not the right one.
- Not Candidate-the ones that are not.
- Others that not:Low plane count-the ones that have trouble fulfilling cuts.
- Failure mode examples
Reporting (Did you find 'something'? What's next?)¶
- Procedures:
Other links and references¶
Recent pdf's in docdb:
- FD Neutrino Scan Procedure (J. Musser) docdb-10372
Miscellaneous¶
Any other useful stuff.