__group__ ticket summary component version type owner status created _changetime _description _reporter 17 Add W-tagging tools JetTools 4.0 enhancement cvermilion reopened 2011-01-03T16:22:43Z 2011-07-29T20:53:22+01:00 "Add as many as possible of the tools in W-tagging paper of Schwartz et al., especially R-cores. W-tagging, as implemented in the Schwartz paper, is now available, but just R-cores would be nice, too. (Are there other new ones?)" cvermilion 51 Allow no input particles in GUI display 4.0 defect cvermilion new 2011-11-10T18:55:38Z 2011-11-10T18:55:38Z You can run !SpartyJet without storing input particles; you should be able to use the GUI on the resulting output files. Right now trying to run event displays segfaults when you do this. Clean up SpartyGUI.py so that we test if inputs are stored where necessary. cvermilion 32 b tagging IO 4.0 enhancement cvermilion new 2011-02-08T16:23:39Z 2011-02-08T16:23:48Z "Some kind of crude b-tagging. This really requires looking back through truth data, so it depends on the input type! ""Pretty good"" approach would be to find B hadron, and then if it's within R of a jet axis, tag the jet. ""OK"" would be to do this with the b quark (presumably some output files will store this but not the B hadron)." cvermilion 23 Beam thrust moment tool JetTools 4.0 enhancement cvermilion assigned 2011-01-24T21:59:20Z 2011-07-29T20:54:18+01:00 Add a JetMomentTool implementing beam thrust (see talk by Neubert at Boston Jet Workshop, etc.). cvermilion 59 Bug in Pythia input? IO 4.0 defect cvermilion new 2013-04-25T03:13:54+01:00 2013-04-25T03:13:54+01:00 "Per a user email: I downloaded spartyjet-pythia from http://www.pa.msu.edu/~marti347/project6/index.html. I believe I might have found a bug in the way pythia files are read. In JetCore/PythiaInput.cc, final state particles are identified using (status == 1), in pythia this should be done using (status > 0.) " cvermilion 21 Check phi conventions misc 4.0 task cvermilion assigned 2011-01-12T23:20:59Z 2011-05-15T21:55:57+01:00 PseudoJet::phi() returns [0,2pi], but I think our old convention was [-pi,pi] (see recent versions of EventView.cc, eg). We should change phi() to phi_std() everywhere that it matters! Potentially a big bug somewhere, maybe not. cvermilion 40 Clarify Moment storage JetCore 4.0 task cvermilion assigned 2011-05-15T20:19:24+01:00 2012-03-28T19:24:11+01:00 The way !JetMomentMaps get passed around now seems somewhat confusing. Do they go with the !JetCollection or with the !JetAnalysis? cvermilion 14 Clean up (restructure?) SpartyGui display 4.0 cleanup cvermilion new 2010-11-17T16:53:21Z 2010-11-17T16:53:21Z "SpartyGui.py is pretty bloated. It might be nice to class-ify some of its features and make the whole ""JetCollectionItems"" concept clearer." cvermilion 5 Cleanup JetCollection interface JetCore 4.0 cleanup cvermilion assigned 2010-11-17T16:26:23Z 2011-05-19T03:22:36+01:00 The !JetCollection interface has gotten really bloated (my fault!). We should trim back what it does, and make the big picture sensible. The goal should be to bring !JetCollection and !ClusterSequence as close together as possible while retaining easy subjet modification. cvermilion 41 Code cleanup misc 4.0 cleanup cvermilion assigned 2011-05-15T22:41:57+01:00 2011-05-15T22:42:06+01:00 The codebase is a mess; spacing, underscores vs. !CamelCase, weird naming, etc. Go through the whole thing at once and clean up! (There are tools for this, no?) cvermilion 56 Convert all cut tools to use Selectors JetTools 5.0 cleanup cvermilion new 2012-03-28T19:22:41+01:00 2012-03-28T19:22:41+01:00 !FastJet Selectors should deprecate our various cut tools. cvermilion 48 Create PyInputMaker python 4.0 enhancement cvermilion assigned 2011-09-07T17:20:29+01:00 2011-09-07T17:21:05+01:00 Want an analogue to !FourVecInput for use in Python scripts. !PyFourVecInput : !FourVecInput :: TPySelector : TSelector (in PROOF). Basically, need a C++ class that expects to be called from Python. Should anticipate that a Python class will derive from it, so do the !PyObject* function calls, a la TPySelector. cvermilion 47 Ditau tagging JetTools 4.0 task cvermilion assigned 2011-08-17T19:45:12+01:00 2011-08-29T16:06:22+01:00 Add a di-tau tagger, based on Tuhin's paper or Brock's (haven't looked if they're the same). Suggested by Tuhin, who sent me code for charged particle ID. cvermilion 50 Document setup better doc 4.0 enhancement cvermilion assigned 2011-09-21T17:19:33+01:00 2011-09-21T17:41:20+01:00 "Several features of getting SJ running could be better documented/done. At minimum: * To use Python features, ROOT must be built with PyROOT enabled, and ROOTSYS/lib must be in your PYTHONPATH. * !FastJet is currently being linked dynamically, which means that FASTJET/lib must be in (DY)LD_LIBRARY_PATH. * Right now the build requires that you have configured !FastJet to include !SISCone, but this is not !FastJet's default configuration. Only use !SISCone if the user has it!" cvermilion 28 "Enhance ""on-the-fly"" to include all JetTools" display 4.0 enhancement cvermilion new 2011-01-24T22:07:11Z 2011-01-24T22:07:11Z """On-the-fly"" algorithms would be much more useful if the whole tool chain were manipulable from the GUI. Drop-down menus?" cvermilion 52 Fault tolerance JetCore 4.0 defect cvermilion assigned 2011-12-10T18:26:15Z 2011-12-10T18:26:23Z Lots of SJ components crash/segfault when bad input is given (e.g., input makers with invalid files names). Clean this up and fail gracefully! cvermilion 19 Find input name in SpartyGui display 4.0 enhancement cvermilion new 2011-01-03T17:20:38Z 2011-01-03T17:20:38Z SpartyGui should be able to figure out the name of the input branch in the ROOT file -- it is normally the one that contains 'Input'. Make specifying this optional. cvermilion 54 Function naming and argument ordering cleanup JetCore 5.0 task cvermilion new 2012-03-28T19:20:26+01:00 2012-03-28T19:20:26+01:00 "Current function naming and argument ordering ""conventions"" are a complete mess. This should really be cleaned up! Function names should probably be underscore_separated() to match !FastJet. For tools, there should be some convention about whether the name goes at the beginning or the end of the argument list, etc." cvermilion 27 GUI interface to Pythia? JetTools 4.0 enhancement cvermilion new 2011-01-24T22:06:02Z 2011-01-24T22:06:02Z Add a GUI interface to Pythia event generation. This would make the whole analysis chain accessible from the GUI -- totally code-free !SpartyJet! cvermilion 35 Identify tagged jets in event views JetTools 4.0 enhancement cvermilion new 2011-02-08T16:28:18Z 2011-02-08T16:28:18Z """Taggers"" should probably just store a bool moment for jets. It would then be nice to be able to use this later in the event views, eg, to only color tagged jets." cvermilion 36 Implement cuts in event views display 4.0 enhancement cvermilion new 2011-02-08T16:30:19Z 2011-02-08T16:30:19Z If a TCut is set in the cut dialog box, the event viewer should just show jets that pass the cut. There should be an option to only show events with at least one jet passing. cvermilion 29 Implement Kribs/Martin/Roy t' search examples 4.0 enhancement cvermilion new 2011-02-08T16:17:01Z 2011-02-08T16:17:01Z This search uses the HEP top tagger, the BDRS Higgs tagger, and several other tools, so would be a nice example. Requires some sort of lepton isolation and b-tagging. cvermilion 42 Input reading for display display 4.0 cleanup cvermilion new 2011-05-23T13:36:35+01:00 2011-05-23T13:36:35+01:00 Clean up the input reading for the GUI (!JetCollReader and !JetRetrieverFile). Seems like this is duplicating some !JetCore code. Also, inputs collections don't have the *_ind branch, so we get an annoying warning every time we launch the GUI. Use a !FooCollection, not !FooCollection_index, for the input collection. cvermilion 18 InputMakerFactory python 4.0 enhancement cvermilion new 2011-01-03T17:18:58Z 2011-01-03T17:19:13Z Take InputMaker helper functions in !SpartyConfig.py and put them in a class. cvermilion 43 JetMoment > JetTools 4.0 defect cvermilion assigned 2011-05-24T16:20:01+01:00 2011-07-29T20:47:32+01:00 Storage of zcutMoment doesn't seem to work, presumably related to ROOT choking on the vector. For some reason almost all stored moments only appear to have two elements. (GUI gives tons of errors when trying to plot $$_zcut[2].) cvermilion 31 Lepton isolation JetTools 4.0 enhancement cvermilion new 2011-02-08T16:21:11Z 2011-02-08T16:21:11Z We should implement some sort of crude lepton isolation -- eg, remove any lepton with pT > x and with sum(E_i) < y for all particles i within DR < z. cvermilion 7 Make sure all tools/plugins are documented doc 4.0 enhancement cvermilion assigned 2010-11-17T16:34:11Z 2011-01-24T21:55:51Z "Check the manual to make sure all !JetTools and external plugins are documented. As many as feasible should be included in the examples. This was largely done for 3.5; I'm moving to 4.0 to remind us to document anything we add before then!" cvermilion 26 More energy flow/moment tools? JetTools 4.0 enhancement cvermilion new 2011-01-24T22:04:49Z 2011-01-24T22:04:49Z Inspired by J. Gallicchio's talk at Boston Jet Workshop. He has code for all his variables, may be able to just wrap very simply. cvermilion 24 N-jettiness JetTools 4.0 enhancement cvermilion assigned 2011-01-24T22:00:05Z 2011-05-31T19:58:54+01:00 Add an N-jettiness event moment. Should be very similar to N-subjettiness, which has already been implemented. cvermilion 55 New set of examples examples 5.0 enhancement cvermilion new 2012-03-28T19:21:52+01:00 2012-03-28T19:21:52+01:00 "A new set of examples should at minimum encompass the following: * New top tagging tools from !FastJet * the Boost analyses you can grab easily * !FastJet Selectors * probably other new features I'm forgetting..." cvermilion 30 Pile-up tagging IO 4.0 enhancement cvermilion new 2011-02-08T16:19:33Z 2011-02-08T16:19:33Z If we add pile-up, it would be nice to keep track of those particles to see, for example, what their contribution is. This could be implemented with some sort of input particle tagging (bool moments?). cvermilion 53 Remove internal history JetCore 5.0 cleanup cvermilion new 2012-03-28T19:18:35+01:00 2012-03-28T19:18:35+01:00 Now that !FastJet's !PseudoJet has full-featured substructure, there is no need to carry around the SJ version as well. Time to clear out the internal cruft. For tools that unavoidably need a different structure, we can convert then. We may convert to something like a tree of HistoryElement's for storage, but I don't think it's necessary to have a parallel jet structure around by default. cvermilion 12 Revise Python wrapping python 5.0 enhancement cvermilion assigned 2010-11-17T16:47:56Z 2011-02-08T16:14:02Z "We should consider using SWIG instead of PyROOT as the Python wrapper to !FastJet and !SpartyJet internals. The ROOT dictionary system is pretty ugly, and not everyone wants to deal with ROOT (just having a Python interface to !FastJet would be nice!). Also, classes that won't end up being store in a file don't need ROOT dictionaries. (Anything seen by the user in a Python script needs either a PyROOT or SWIG wrapper.) " cvermilion 49 Rework jet moments JetCore 4.0 task cvermilion assigned 2011-09-10T14:51:36+01:00 2011-09-10T14:51:42+01:00 "Instead of a !JetMomentMap that lives with a !JetCollection, jet moments should be associated with jets. The current approach makes it difficult to write tools in terms of jets, because you need access to the JMM to set moments. Thus basically every jet-based tool has the exact same few lines of code that does the loop. There's got to be a better way! Possibly solved by moving toward full !PseudoJet behavior for Jets? !JetCollection as a !PseudoJetStructureBase?" cvermilion 60 ROOT compatibility JetCore 4.0 defect cvermilion new 2013-04-29T04:19:29+01:00 2013-04-29T04:19:29+01:00 A loyal user points out that some of the definitions in JetCore/JetCoreLinkDef.hpp break in recent ROOT versions. We should take a look at what versions of ROOT we want to support and check for deprecation. cvermilion 37 Save substructure for taggers? JetTools 4.0 enhancement cvermilion new 2011-02-08T16:33:35Z 2011-02-08T16:33:35Z Can we store (do we store?) constituent and subjet information for taggers whenever possible? This would be useful, for example, in the event displays. cvermilion 6 Storable JetHistory JetCore 4.0 enhancement cvermilion assigned 2010-11-17T16:28:20Z 2010-11-17T16:55:51Z Allow !JetHistory to be stored in a ROOT file. This can then be used by the GUI. I believe a working version is in the spartyjet_newhistory branch, but this branch is very outdated, so merging will probably be tricky. cvermilion 22 store record of tool chain misc 5.0 enhancement cvermilion assigned 2011-01-14T21:12:28Z 2011-05-19T03:24:50+01:00 It would be great if the output ROOT file included the original script that generated it! (Requested by Sal Rappoccio.) cvermilion 25 Template overlap JetTools 4.0 enhancement cvermilion new 2011-01-24T22:03:14Z 2011-01-24T22:03:14Z "Is there some way we can implement ""template overlap"" methods in SJ, a la the top-finding technique of Gilad Perez, et al.? Look into this; they claim to be working on code..." cvermilion 58 vector> ntuple input IO 5.0 enhancement cvermilion new 2012-08-17T14:14:16+01:00 2012-08-17T14:14:16+01:00 "Request from Anna Sfyria of ATLAS: I have been checking out past presentations of yours and see you accept requests? :-) If that's the case, I would like to suggest to add an input reader of vectors of vectors of floats, rather than just vectors of floats. The reason is that often the jet constituents to be used as input in your tools are saved in vector> branches on the ntuples (D3PDs) used in the physics groups on ATLAS. Having an ntupleInputMaker that can read them directly would be super useful! " cvermilion 9 Working history display display 4.0 defect cvermilion new 2010-11-17T16:38:19Z 2010-11-17T16:38:19Z Current history !GUI is pretty buggy. Cleaning this up should probably wait for the storable !JetHistory, to make sure that interface doesn't need to change. cvermilion