Les Houches

User Tools

Site Tools

Trace:
2015:groups:tools:lhaad

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
Last revision Both sides next revision
2015:groups:tools:lhaad [2015/06/18 12:25]
daniel.schmeier 		2015:groups:tools:lhaad [2016/02/25 17:41]
philippe.gras
Line 1: Line 1:
 ===== Les Houches Analysis Description Accord ===== ===== Les Houches Analysis Description Accord =====
  
-People: S. Sekmen, H. Prosper, G. Unel, B. Fuks, J. Tattersall, D. Schmeier, C. Pollard, S. Kraml, S. Kulkarni, D. Morse, D. Majumder, N. Desai, D. Sengupta. N. De Filippis, J. Sonneveld ... etc. (add your name if interested)+People: S. Sekmen, H. Prosper, G. Unel, B. Fuks, J. Tattersall, D. Schmeier, C. Pollard, S. Kraml, S. Kulkarni, D. Morse, D. Majumder, N. Desai, D. Sengupta. N. De Filippis, J. Sonneveld, E. Conte, Ph. Gras ... etc. (add your name if interested) 
 + 
 +//If you are interested in contributing please subscribe to the [[https://​e-groups.cern.ch/​e-groups/​EgroupsSearch.do?​searchValue=houches-2015-topics-tools-lhada|Mailing list]]. You will be asked to log in with your CERN account. If you don't have a CERN account, please fill [[https://​account.cern.ch/​account/​Externals/​RegisterAccount.aspx|this form]] to create a lightweight account.//
  
 Goal: Prepare an SLHA-like easy to read and write accord for describing the flow of an LHC analysis. Goal: Prepare an SLHA-like easy to read and write accord for describing the flow of an LHC analysis.
Line 7: Line 9:
 First proposal made for a generic structure. ​ CM and MA authors will try to implement a single lepton stop search. First proposal made for a generic structure. ​ CM and MA authors will try to implement a single lepton stop search.
  
-Proposal by Daniel Schmeier & Jamie Tattersall +Proposal by Daniel Schmeier & Jamie Tattersall ​[[2015:groups:tools:lhaad:proposal1]]
- +
-<code autoconf>​ +
-# These are the input objects for the analysis code. +
-# Syntax:  +
-#      name    type (see below) +
-# A well formulated comment behind each object should unambiguously state what this object is +
-# such that a _human_ knows what this object corresponds in his/her code +
-       +
-BLOCK ANALYSISINPUT +
-  jets      l<​jet> ​     #   ​p4,​pt>​20 ​   alg=anti-kt ​   R=0.4    b_prob,​eff=(60,​70,​80) ​   +
-  elecs     ​l<​el> ​      # ​  ​loose,​ isotags 1=tracks,​dr0.2,​ 2=... +
-  muons     ​l<​mu> ​      # +
-  etmiss ​   v           # +
-   +
-   +
-# All functions that have to be used later must be defined here. +
-# The declaration is similar to typical type-based declaration: It has to be +
-#  specified what the name of the function is, what objects go in and what comes out. +
-# +
-# Syntax:  ​Function_name ​  ​input_type_1 ​ input_type_2 ​  ​input_type_3 ... output_type +
-# Rules: +
-#        - input_types and output_type can be  +
-#           - '​s'​ for scalars/​numbers,​  +
-#           - '​v'​ for LorentzVectors, ​        +
-#           - '​b'​ for booleans +
-#           - '​c'​ for characters/​strings +
-#           - any of the objects defined in the '​object'​ block +
-#           - l<​type>​ to denote a list of objects of common type <​type>​. +
-#        - People will comment with a link to a resource where the exact definition +
-#          of that function can be found, in text definiton and a working example in a programming language of choice +
-#        - 2 functions can have the same name if they have different input types +
- +
-BLOCK ANALYSISFUNCTIONS +
-  mT               ​v ​   v    s             # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​LHA/​Definitions/​m_T +
-  amT2             ​v ​   v    s             # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​LHA/​Definitions/​am_T2 +
-  mT2tau ​          ​v ​   v    s             # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​LHA/​Definitions/​m_T2_tau +
-  mHadTop ​         ?    ?    ?             # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​LHA/​Definitions/​m_had_top +
-  HTMissSig ​       ?    ?    ?             # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​LHA/​Definitions/​HTmisssig +
-   +
-  # s denotes the working point efficiency +
-  filter_no_bjets ​     l<​jet> ​ s      l<​jet> ​  # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​BJetDefinitionSource +
-  filter_bjets ​        ​l<​jet> ​ s      l<​jet> ​  # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​BJetDefinitionSource +
-  filter_taujets ​      ​l<​jet> ​ s      l<​jet> ​  # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​TauJetDefinitionSource +
-  filter_medium_elec ​  ​l<​elec>​ l<​elec> ​        # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​ElectronIDSource +
-  filter_tight_elec ​   l<​elec>​ l<​elec> ​        # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​ElectronIDSource +
-    +
-  #                    input  whichsource ​ dr       ​ptmin ​  ​maxval ​  ​output +
-  filter_iso_abs ​      ​l<​el> ​ s    s   ​s ​  ​s ​  ​l<​el> ​    # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​Definitions/​Isolation +
-  filter_iso_rel ​      ​l<​el> ​ s    s   ​s ​  ​s ​  ​l<​el> ​    # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​Definitions/​Isolation +
-  filter_iso_abs ​      ​l<​mu> ​ s    s   ​s ​  ​s ​  ​l<​mu> ​    # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​Definitions/​Isolation +
-  filter_iso_rel ​      ​l<​mu> ​ s    s   ​s ​  ​s ​  ​l<​mu> ​    # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​Definitions/​Isolation +
-   +
-  trigger ​             l<​el> ​ l<​mu> ​ l<​jet> ​ v           # https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​Definitions/​SingleLepTrigger + EtMissTrigger +
-  dataquality ​         l<​jet>​ v                          # Tuned to a constant efficiency of 99.7%, see here: https://​twiki.cern.ch/​twiki/​bin/​view/​XXPublic/​EventCleaning/ ​  +
-   +
-# These functions are like normal functions, but as these are always needed one might find a  +
-#  way to globally use them without continuously redefining them +
-  sum               ​s ​  ​s ​  s +
-  diff              s   ​s ​  s +
-  prod              s   ​s ​  s +
-  quot              s   ​s ​  s +
-  pow               ​s ​  ​s ​  s +
-  random ​           s  +
-   +
-  overlap_removal ​  ​l<​v> ​  ​l<​v> ​  ​s ​    v +
-  delta_phi ​        ​v ​     v      s +
-  delta_r ​          ​v ​     v      s +
-  ptmin             ​l<​v> ​  ​s ​     l<v>  +
-  etarange ​         l<​v> ​  ​s ​     s     ​l<​v>​ +
-  combine ​          ​l<​v> ​  ​l<​v> ​  ​l<​v>​ +
-  take_leading_n ​   l<​v> ​  ​s ​     l<​v>​ +
-          +
-   +
-   +
-# The Preselection block only trims existing objects or define new objects by applying functions on existing ones. +
-# It might be hard to read at first, but one just has to imagine the column +
-#      x     ​f ​    ​y ​    z +
-# with the following extra characters +
-#      x  =  f  (  y  ,  z  ) +
-# i.e. one associated the object in col1 with the result of the function in col2, putting in the objects that come after. +
-# +
-# Syntax: ​  ​new_object function parameter1 ​  ​parameter2 ​ ... +
-# Rules: +
-#     - each function must be defined in the function block +
-#     - parameters must match the parameter type definition in the function block   +
-#     - parameters must be numbers, strings, objects that have been defined in previous lines or in the '​object'​ block +
-#     - Every TLorentzVector function ​ "​Double_t Fun() const" can be accessed by v.fun (eg. v.m, v.pt, v.px, v.costheta +
-#         +
-#     - With vectors one can do +
-#        vectors.n ​  ​returns size +
-#        vectors.3 ​  ​returns the third(!) object in the list +
-#        vectors.8 ​  ​returns a zero object if vectors contains less than 8 entries +
-#     - Cut-Logic should make sure that the last rule does not produce rubbish (see mT example)  +
-#     - objects and functions can have the same name as the syntax does clearly distinguish between the two +
- +
-BLOCK ANALYSISPRESELECTION +
-  # Kinematic Cuts on the loose baseline and the signal objects +
-  jets        ptmin           ​jets ​       20  +
-  jets        etarange ​       jets        -2.5       2.5 +
-  elecs_l ​    ​ptmin ​          ​elecs ​      ​7 ​   +
-  elecs_l ​    ​etarange ​       elecs_l ​    ​-2.47 ​     2.47   +
-  elecs_m ​    ​filter_elecs_m ​ elecs_l +
-  elecs_m ​    ​ptmin ​          ​elecs_m ​   10   +
-  muons_l ​    ​ptmin ​          ​muons ​     6  +
-  muons_l ​    ​etarange ​       muons      -2.4       2.4 +
-  muons_m ​    ​ptmin ​          ​muons ​     10   +
-   +
-  # overlap removals between bjets, nonbjets and leptons +
-  nob_jets ​       filter_no_bjets ​  ​jets ​       70 +
-  b_jets ​         filter_bjets ​     jets        70 +
-  no_bjets ​       overlap_removal ​  ​no_bjets ​   elecs_m ​  0.2 +
-  elecs_m ​        ​overlap_removal ​  ​elecs_m ​    ​no_bjets ​ 0.4 +
-  elecs_m ​        ​overlap_removal ​  ​elecs_m ​    ​bjets ​    0.2 +
-  elecs_m ​        ​overlap_removal ​  ​elecs_m ​    ​tau_jets ​ 0.2 +
-  muons_m ​        ​overlap_removal ​  ​muons_m ​    ​jets ​     0.4 +
-  tau_jets_l ​     filter_taujets ​   "​loose" ​    ​jets ​   +
-  tau_jets ​       filter_taujets ​   "​tight" ​    ​jets ​   +
-  tau_jets ​       overlap_removal ​  ​tau_jets ​   elecs_l ​ 0.2 +
-   +
-  # choose signal leptons which have to pass isolation conditions +
-  elecs_t ​       ptmin              elecs_m ​  25 +
-  elecs_t ​       filter_tight_elec ​ elecs_t ​   +
-  elecs_t ​       filter_iso_rel ​    "​tracks" ​     0.2    0.5   0.1 +
-  muons_m ​       filter_iso_abs ​    "​tracks" ​     0.2    0.5   1.8 +
-   +
-  # combine electrons and muons to leptons +
-  leptons_l ​     combine ​           muons_l ​      ​muons_l +
-  leptons_l ​     ptmin              leptons_l ​    10 +
-  leptons ​       combine ​           elecs_t ​      ​muons_m +
-  leptons ​       ptmin              leptons ​      25 +
-   +
-  # find the leading 4 jets +
-  leading_jets ​  ​take_leading_n ​    ​jets ​         4 +
-  leading_bjets ​ filter_bjets ​      ​leading_jets ​ 70 +
- +
-  # Calculate some advanced kinematical quantities +
-  HT             ​sum ​        ​jets.1.pt ​  ​jets.2.pt +
-  HT             ​sum ​        ​HT ​         jets.3.pt +
-  HT             ​sum ​        ​HT ​         jets.4.pt +
-  sqrtHT ​        ​sqrt ​       HT +
-  EtoverHT ​      ​quot ​       etmiss.pt ​  ​sqrtHT +
-  dPhi_j1_miss ​  ​dPhi ​       jets.1 ​     etmiss +
-  dPhi_j2_miss ​  ​dPhi ​       jets.2 ​     etmiss +
-  mT             ​mT ​         etmiss ​     leptons.1 +
-  amT2           ​amT2 ​       ?           ? +
-  mHadTop ​       mHadTop ​    ? ​          ? +
-  mT2Tau ​        ​mT2Tau ​     ?           ? +
-  HTMissSig ​     HTMissSig ​  ? ​          ? +
-   +
-  # Find overall efficiency for event +
-  trigeff ​       trigger ​    ​elecs_l ​    ​muons_l ​  ​jets ​  ​etmiss +
-  cleaningeff ​   dataquality elecs_l ​    ​muons_l ​  ​jets ​  ​etmiss ​  +
-  eventeff ​      ​prod ​       trigeff ​    ​cleaningeff ​  +
-  trigrand ​      ​random +
-  .... +
-   +
-# CUT blocks contain combinations of boolean statements +
-# Syntax: +
-#    Variable ​  ​operator ​  ​Value +
-# Rules: +
-#    - operator: ​ <, <=, =, >, >=, != +
-#    - allowed Variables are:  +
-#       - input objects +
-#       - everything defined in the preselection block +
-#       - a preceeding CUT can be condidered as a Boolean +
-#    - functions cannot be used here. If the result of a function is needed, it must be evaluated in the preselection block +
-#    - there can be more than one condition in one CUT. Obviously this means that all conditions have to be fulfilled ​            +
- +
-CUT  common +
-  leptons.n ​  ​= ​  1 +
-  leptons_l.n =   1 +
-  jets.n ​     >=  4 +
-  etmiss.pt ​  > ​  100 +
-   +
-       +
-CUT leadjets_60to25 +
-  jets.1.pt ​  > 60 +
-  jets.2.pt ​  > 60 +
-  jets.3.pt ​  > 40 +
-  jets.4.pt ​  > 25 +
-   +
-CUT leadjets_80to25 +
-  jets.1.pt ​  > 80 +
-  jets.2.pt ​  > 60 +
-  jets.3.pt ​  > 40 +
-  jets.4.pt ​  > 25 +
-   +
-CUT leadjets_100to25 +
-  jets.1.pt ​  > 100 +
-  jets.2.pt ​  > 80 +
-  jets.3.pt ​  > 40 +
-  jets.4.pt ​  > 25 +
-   +
-CUT large-R-jet +
-  fat_jets.n >= 1 +
-  fat_jets.1.m > 75 +
-   +
-CUT dPhi12 +
-  fat_jets.n >= 2   +
-  dPhi_j1_miss ​  > 0.8 +
-  dPhi_j2_miss ​  > 0.8 +
-   +
-CUT tN_diag +
-  common ​         =  True +
-  leadjets_60to25 =  True +
-  leading_bjets.n >= 1 +
-  mT              > 100 +
-  mHadTop ​        > 130 +
-  mHadTop ​        < 205 +
-  taujets.n ​      = 0 +
-  dRbl            < 2.5 +
-  EtoverHT ​       > 5 +
-  dPhi12 ​         = True +
-   +
-  +
-CUT tN_med +
-  common ​         =  True +
-  leadjets_80to25 =  True +
-  leading_bjets.n >= 1 +
-  etmiss.pt ​      > 200 +
-  mT              > 140 +
-  amT2            > 170 +
-  mHadTop ​        > 130 +
-  mHadTop ​        < 195 +
-  HTMissSig ​      > 12.5 +
-  dPhi12 ​         = True +
-   +
-   +
-CUT tN_high +
-  common ​         =  True +
-  leadjets_80to25 =  True +
-  leading_bjets.n >= 1 +
-  etmiss.pt ​      > 320 +
-  mT              > 200 +
-  amT2            > 170 +
-  mT2tau ​         > 120 +
-  mHadTop ​        > 130 +
-  mHadTop ​        < 250 +
-  dRbl            < 3.0 +
-  HtMissSig ​      > 12.5 +
- +
-# Cuts are interpreted as "​remove event if it fails this". However, some conditions might just +
-# apply a weight to the event. These can be stated in the following block +
-# This can also be used for general event re-weighting e.g in the case of jet matching +
-# Syntax: +
-#      Cut   ​weight +
-# Rules: +
-#      - Cut must correspond to a CUT block +
-#      - weight must be a scalar, can be a variable though+
  
-BLOCK WEIGHTS +Proposal by Nishita Desai, Suchita Kulkarni & Devdatta Majumder [[2015:groups:​tools:​lhaad:​proposal2]]
-  common ​  ​eventeff +
-   +
-      +
-   +
-# BINNING blocks simplify exclusive bin-cuts and easily identify orthogonal regios +
-# Syntax: +
-#    BINNING (somename) +
-#      Cut        (name of cut after which binning is applied) +
-#      Parameters ​ (parameter1 to bin),(parameter2 to bin),... +
-#      New-Cutname1 ​  ​bin1_min1,​bin1_min2,​... ​  ​bin1_max1,​bin1_max2,​... +
-#      New-Cutname2 ​  ​bin2_min1,​bin2_min2,​... ​  ​bin2_max1,​bin2_max2,​... +
-#      ... +
-# Rules: +
-#    - Cut and Parameter must be valid, see above +
-#    - New-Cutname1 is a new Cut just like it was defined via a "CUT " block +
-#    - -inf and inf for "​everything less/more than"​ +
-#    - Bins don't have to be orthogonal. Still it is easy to automatically check orthogonality a posteriori +
-    ​+
  
-BINNING tN_diag_mTEtMiss +Combined Proposal by ND, SK, DM, DS & JT: [[2015:​groups:​tools:​lhaad:​proposal]]
-  Cut       ​tN_diag_g100 +
-  Parameter mT,etmiss +
-  tN_diag_g100_g60 ​  100,60   125,90 +
-  tN_diag_g100_g90 ​  ​100,​90 ​  ​125,​120 +
-  tN_diag_g100_g120 ​ 100,​120 ​ 125,150 +
-  tN_diag_g100_g140 ​ 100,​150 ​ 125,inf +
-  tN_diag_g125_g60 ​  ​125,​60 ​  ​150,​90 +
-  tN_diag_g125_g90 ​  ​125,​90 ​  ​150,​120 +
-  tN_diag_g125_g120 ​ 125,​120 ​ 150,150 +
-  tN_diag_g125_g140 ​ 125,​150 ​ 150,inf +
-  tN_diag_g150_g60 ​  ​150,​60 ​  ​inf,​90 +
-  tN_diag_g150_g90 ​  ​150,​90 ​  ​inf,​120 +
-  tN_diag_g150_g120 ​ 150,​120 ​ inf,150 +
-  tN_diag_g150_g140 ​ 150,​150 ​ inf,inf+
  
 +Proposal to use HEPDATA format for the experiment data section. HEPDATA (https://​hepdata.net/​) format has recently changed, both old and new formats are text file. New format: https://​github.com/​HEPData/​hepdata-submission (Old format: http://​hepdata.cedar.ac.uk/​submittingdata , see "​simple input file"​).
  
-# Experimental Results are written in the following block 
-# Syntax: 
-#    SRname ​ Obs   ​Bkg ​  ​Bkgerr1 ​  ​Bkgerr2 ​ Bkgerr3 ... 
-# Rules: 
-#    Each SRname must correspond to a cut from above 
-BLOCK ExpResults: 
-  tN_diag_g100_g60 ​  ​3452 ​  ​3452 ​  59 
-  tN_diag_g100_g90 ​  ​1712 ​  ​1720 ​  ​161 ​   ​ 
-  tN_diag_g100_g120 ​ 313    295    50 
-  tN_diag_g100_g140 ​ 201    235    34 
-  tN_diag_g125_g60 ​  ​2018 ​  ​2018 ​  ​45 ​ 
-  tN_diag_g125_g90 ​  ​768 ​   767    80 
-  tN_diag_g125_g120 ​ 117    136    22 
-  tN_diag_g125_g140 ​ 163    152    30 
-  tN_diag_g150_g60 ​  ​2543 ​  ​2543 ​  ​50  ​ 
-  tN_diag_g150_g90 ​  ​647 ​   684    79 
-  tN_diag_g150_g120 ​ 98     ​13 ​    101 
-  tN_diag_g150_g140 ​ 217    236    29 
-  tN_med ​            ​12 ​    ​13.0 ​  2.2 
-  tN_high ​           5      5.0    1.0 
  
-# Cutflows show events after consecutive evaluation of conditions +Agreements: 
-# Syntax: +  ​* Binned information should have a simple way to cut it 
-#    Parameter Operator Value   ​EventsAfterCut1 ​  ​Uncertainty ​  "​Description in LateX Format"​ +  ​* Cuts must not branch but instead can use a common '​predecessor'​ 
-# Rules: +  ​* Syntax must not contain brackets to allow for an easy parsing 
-#    - Parameter operator Value combinations:​ See above +  ​* ...
-#    - Cuts are applied one after another +
-#    -  +
-    +
-CUT mHadTopRange +
-    mHadTop ​        > 130 +
-    mHadTop ​        < 205    +
-     +
-CUTFLOW ​ tN_diag_echannel +
-    ​True ​            ​= ​ True     ​200000 ​  ​0 ​  "​Initial Events"​ +
-    ​trigrand ​        < ​ trigeff ​ 113762 ​  ? ​  "​Trigger"​ +
-    ​leptons_l.n ​     =  1        28785    ?   "​Lepton (exactly 1 baseline)"​ +
-    leptons.n        =  1        19875    ?   "​Lepton (exactly 1 signal)"​ +
-    jets.n          >=  3        17959    ?   "​At least 3 jets $\geq$ 25 GeV" +
-    jets.n          >=  4        13682    ?   "​At least 4 jets $\geq$ 25 GeV" +
-    jets.1.pt ​      >​= ​ 60       ​12804 ​   ?   "​jet1pt $\geq$ 60" +
-    jets.2.pt ​      >​= ​ 60       ​9489 ​    ? ​  "​jet2pt $\geq$ 60" +
-    jets.3.pt ​      >​= ​ 40       ​8556 ​    ? ​  "​jet3pt $\geq$ 40" +
-    leading_bjets.n >=  1        7192     ? ​  "​btag(70) >= 1" +
-    dPhi_j1_miss ​   >=  0.8      6611     ? ​  "​dPhi(jet1,​ etmiss) >= 0.8" +
-    dPhi_j1_miss ​   >=  0.8      5147     ? ​  "​dPhi(jet2,​ etmiss) >= 0.8" +
-    etmiss.pt ​      >​= ​ 150      1314     ? ​  "​Etmiss >= 150" +
-    EtoverHT ​       >=  5        1314     ? ​  "​Etmiss/​sqrtHT >= 5" +
-    mT              >=  140      446      ?   "​mT >= 140 GeV" +
-    mHadTopRange ​   =   ​True ​    ​283 ​     ?   "​mHadTop in 130, 205" +
-    taujets_l.n ​    ​= ​  ​0 ​       257      ?   "​veto on loose Taus"​ +
-</​code> ​    +
-    ​+
  
 +To Be Discussed Isues:  ​
 +  * Format
 +  * Definition of '​detector constructed'​ objects like jets, electrons, etc
 +  * How to enforce documentation of the used objects
 +  * How to enforce documentation of validation material
 +  * ...
  
 +Suggestion:
 +  * Acronym: LHADA (Les Houches Analysis Description Accord) and not LHAAD
 +  * Information about benchmark points
 +  * Information about process generation needs to be added (merging parameters, generated process, link SLHA file if needed, link to model files etc.)
2015/groups/tools/lhaad.txt · Last modified: 2016/02/26 10:18 by daniel.schmeier

Page Tools

Except where otherwise noted, content on this wiki is licensed under the following license: Public Domain
Public Domain Donate Powered by PHP Valid HTML5 Valid CSS Driven by DokuWiki