Skip to content
Snippets Groups Projects
Commit daf9b987 authored by Klaus Rabbertz's avatar Klaus Rabbertz
Browse files

Fix steering filename and scenarioname inside

parent 4ca9ed26
Branches master
No related tags found
No related merge requests found
# -*-sh-*-
# ==================================================================== #
#
# A steering file for creating a fastNLO table/grid (with NNLOJET)
# This is an example for a global steering for ALL tables/grids of a scenario
#
# The grid production is performed in three steps:
# I. A warmup run for NNLOJET to perform essential integrations
# See NNLOJET for details. fastNLO is not used or called at this stage.
# II. A warmup run for fastNLO that determines the accessed phase space
# in momentum fraction x and process scale mu_r, mu_f.
# Only a few fastNLO settings are used here, see below.
# The output is written into a warmup file named after the
# the NNLOJET process name, e.g. 'ZJ', the NNLOJET job name, e.g.
# 'LO-CMS13-ak07', and the NNLOJET grid name, e.g.
# ptz > ZJtriple_yb0_ystar0_ptz [30, 100, 200, 300, 400, 1000] grid = ZJtriple_yb0_ystar0_ptz.fast
# NNLOJET histo fastNLO grid
# ==> ZJ.LO-CMS13-ak07.ZJtriple_yb0_ystar0_ptz.wrm
# This warmup file stores some settings for fastNLO in addition to
# the accessed phase space (with number of events per bin!) for
# each order of NNLOJET run (LO, R, V, RRa, RRb, RV, VV).
# For a later combination ALL production runs in step III below
# MUST use the SAME warmup file. The latter should be the largest
# accessed phase space as combined from any order in NNLOJET!
# ==> the expected filename for this unique warmup file is
# ZJ.ZJtriple_yb0_ystar0_ptz.wrm
# without the job name info.
# III. Many production runs filling fastNLO grids for various orders
# of a process from NNLOJET that later must be combined.
# See NNLOJET for details.
# Most fastNLO settings are taken into account here.
#
# This steering file has to be used in combination with a NNLOJET
# run card, where most of the essential seetings like the
# observable and its binning are defined.
# Modifications or additional settings respect the following hierarchy:
# 1. From the interface package nnlo-bridge default settings for
# fastNLO with NNLOJET are made and information from NNLOJET is
# passed to fastNLO.
# 2. A steering file named after the NNLOJET grid name for fastNLO,
# e.g. from the example above
# ==> ZJ.ZJtriple_yb0_ystar0_ptz.str
# can be used to set steer parameters individually for each fastNLO grid.
# Settings from step 1 can be overwritten. Be careful not to make
# changes inconsistent with what is defined in the NNLOJET run card
# or what is mandatory for the interface to work!
# 3a. In production runs the previously derived warmup file is read.
# 3b. A global steering file named after only the first part of
# the NNLOJET grid name for fastNLO, 'ZJtriple', which is also used
# as fastNLO scenario name,
# ==> ZJ.ZJtriple.str
# can be used to set steer parameters globally for all fastNLO grids.
# Settings from step 2 are not overwritten, but lead to warnings.
# 4. In principal, parameters can also be set on the command line, but
# this has not been tested yet.
#
# Special remarks on scale settings:
# ----------------------------------
# Lacking sufficient scale choice information from NNLOJET the following
# settings are assumed for flexible-scale tables that allow two central
# scale choices to be stored simultaneously:
# DIS pp
# - the 1st scale set for muf defines scale1 scale2
# - the 1st scale set for mur defines scale2 scale1
# To check the closure between NNLOJET and fastNLO the proper
# ASYMMETRIC central scale choice must be used when evaluating a
# fastNLO grid with fnlo-tk-cppread, either scale12 or scale21.
#
# It is recommended to set ScaleDescriptionScale1 and
# ScaleDescriptionScale2 accordingly.
#
# If multiple entries per event are requested using NNLOJET COMPOSITE
# histos AND a scale definition varying with each entry is desired,
# then one central scale, e.g. ptmax=ptj1, must be defined for both
# muf and mur, and the InclusiveJets flag must be set in the fastNLO
# steering file.
# In that case the following definitions hold depending on the
# nnlo-bridge code version:
#
# BRIDGE 0.0.40:
# (requires additional fixed-scale combinations of
# muf = 2.718281828459045 mur = 2.718281828459045
# muf = 4.481689070338065 mur = 4.481689070338065
# muf = 4.481689070338065 mur = 2.718281828459045
# muf = 2.718281828459045 mur = 4.481689070338065
# muf = 12.18249396070347 mur = 2.718281828459045
# muf = 2.718281828459045 mur = 12.18249396070347
# in the SCALES section of the NNLOJET run card.)
#
# DIS pp
# - the 1st scale set for muf&mur defines scale1 scale2
# - the stored observable defines scale2 scale1
# To check the closure between NNLOJET and fastNLO the proper
# SYMMETRIC central scale choice must be used when evaluating a
# fastNLO grid with fnlo-tk-cppread, either scale1 or scale2.
#
# BRIDGE 0.0.46:
# (requires additional fixed-scale combinations of
# muf = 90.0171313005 mur = 90.0171313005
# muf = 54.5981500331 mur = 54.5981500331
# muf = 148.4131591026 mur = 148.4131591026
# muf = 54.5981500331 mur = 90.0171313005
# muf = 90.0171313005 mur = 54.5981500331
# muf = 148.4131591026 mur = 90.0171313005
# in the SCALES section of the NNLOJET run card.)
#
# DIS(?) pp
# - the 1st scale set for muf&mur defines scale1 scale2=mu/obs(!)
# - the stored observable defines scale2 scale1
# To check the closure between NNLOJET and fastNLO the proper
# SYMMETRIC central scale choice must be used when evaluating a
# fastNLO grid with fnlo-tk-cppread, i.e. for pp scale1*scale2.
#
# ==================================================================== #
# ==================================================================== #
#
# Scenario specific settings
#
# Either set individually for each fastNLO grid or globally for all
#
# ==================================================================== #
# -------------------------------------------------------------------- #
# Settings to be modified already for Warmup
# (These are stored in the warmup file!)
# -------------------------------------------------------------------- #
#CheckScaleLimitsAgainstBins true # (def.=true) Set limits for scale nodes to bin borders, if possible. Good if scale equals the binned observable.
# Scales and scale factors must be set in the NNLOJET run card
ScaleDescriptionScale1 "m12_[GeV]" # (def.='scale1') Reset the 1st scale name and unit, e.g. "<pT_1,2>_[GeV]" (Note: The 1st scale must always be in units of [GeV]!)
ScaleDescriptionScale2 "m12_[GeV]" # (def.='scale2') Reset the 2nd scale name and unit (ONLY for flexible-scale tables)
#DifferentialDimension 1 # (must be 1) So far the interface to NNLOJET supports only 1-dim histograms. DO NOT CHANGE!
DimensionLabels { # Labels (symbol and unit) for the measurement dimension (from outer to inner "loop")
"m12_[GeV]" # The default following the example above would be: "ptz"
}
#DimensionIsDifferential { # (must be 2) Specify for each dimension whether
# 2 # 0 : the cross section is NOT differential, i.e. there are two bin borders,
#} # but NO division (normalization) by bin width
# 1 : the cross section is point-wise differential, i.e. only one point is given
# 2 : the cross section is bin-wise differential, i.e. there are two bin borders
# Since NNLOJET provides bin-wise differential distributions use option 2. DO NOT CHANGE!
# -------------------------------------------------------------------- #
# Description of scenario
#
# (Write each line of description in quotation marks)
# -------------------------------------------------------------------- #
ScenarioName fnl2412eff-fc-v2 # (def.=NNLOJET process name) No white space allowed here!
# Here: Either one description for all
ScenarioDescription { # Replace generic one-line text by desired (multi-line) description
"d2sigma-dijet_dm12_dym_[pb_GeV]"
"CMS_Collaboration"
"Dijet_m12_ym0-ym4"
"anti-kT_R=0.7"
"CMS-PAPER-QCD-11-004, arXiv:1212.6660, Phys. Rev. D 87 (2013) 112002."
"RIVET_ID=CMS_2013_I1208923/D06-x01-y01,SPIRES_ID=9901922,INSPIRE_RECORD=1208923"
"THEORY_LABEL=NNLOJET modules2 FC"
"provided by:"
"fastNLO_2.5.1"
"If you use this table, please cite:"
" D. Britzger, K. Rabbertz, F. Stober, M. Wobisch, Proc. DIS 2012, 217 (2012), arXiv:1208.3641."
}
PublicationUnits 12 # (def.=12) Cross section unit to be used for output (negative power of 10, e.g. 12->pb, 15->fb)
# For comparison, e.g. via Rivet, should correspond to published data cross sections
# -------------------------------------------------------------------- #
# Cuts
# -------------------------------------------------------------------- #
# To bet set via NNLOJET run card
# -------------------------------------------------------------------- #
# Observable normalization
# -------------------------------------------------------------------- #
#CalculateBinSize true # (def.=true) Calculate bin width from lower and upper bin boundaries
#BinSizeFactor 1. # (def.=1.) Possibility to provide additional normalization factor, e.g. of 2. for bins in |y|
#BinSize { x1 x2 x3 ... } # (def. not set) If 'CalculateBinSize' is 'false' provide table with bin widths 'by hand' for normalization
# -------------------------------------------------------------------- #
# Observable binning
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
# Use either 'SingleDifferentialBinning' or
# 'DoubleDifferentialBinning' or
# 'TripleDifferentialBinning'
# in accord with 'DifferentialDimension' above
# -------------------------------------------------------------------- #
# To bet set via the NNLOJET run card
# ==================================================================== #
#
# Process specific and therefore at least partially theory-code
# dependent settings
#
# ==================================================================== #
# To bet set directly from NNLOJET interface and via the NNLOJET run card
# ==================================================================== #
#
# A few special settings only used with NNLOJET so far
#
# ==================================================================== #
#InclusiveJets true # (def.=false) Flag to vary the mur, muf scale per entry, not just per event
# # Must be set in accordance with NNLOJET run card!
# # Only possible with flexible-scale grids
# ==================================================================== #
#
# General fastNLO settings
#
# ==================================================================== #
GlobalVerbosity INFO # (def.=WARNING) Global output verbosity of fastNLO toolkit,
# (and of user scenario code if implemented).
# Possible values are: DEBUG, MANUAL, INFO, WARNING, ERROR, SILENT
FlexibleScaleTable true # (def.=false) Create table fully flexible in mu_f (larger size), true, or
# table with fixed number of mu_f scale factors
CacheType 20 # For now switch off caching feature
CacheMax 30
CacheCompare 10
# ==================================================================== #
#
# Specific fastNLO settings (normally do not need to be changed)
# Only change when you know what you are doing ...!
#
# ==================================================================== #
#OutputFilename fastNLO.tab # Overwrites default filename of fastNLO output table, e.g. ZJ.LO-CMS13.vBa.ZJtriple_yb0_ystar0_ptz.s92394.tab.gz
#OutputPrecision 8 # (def.=8) Number of decimal digits to store in output table
#OutputCompression true # (def.=true) Write out fastNLO table/grid in gzipped format (requires zlib)
#CacheType 0
#CacheMax 30
#CacheCompare 10
#ScaleVariationFactors { # (def.=1.0 0.5 2.0) Must be set in accordance with NNLOJET run card! Factorization scale variations (only needed for fixed-scale tables)
# 1.0 0.5 2.0 # List of scale factors must include factor '1.0'
#} # Scale factors will be ordered according to fastNLO convention: (1, min, ... , max)
#ReadBinningFromSteering false # (must be true) Specify where the binning is defined.
#ApplyPDFReweighting true # (def.=true) Apply reweighting of PDFs for an optimized interpolation
# -------------------------------------------------------------------- #
# Choose fastNLO interpolation kernels and distance measures
# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
# Currently implemented interpolation kernels
# Catmull
# Lagrange
# OneNode
# Linear
#
# Currently implemented distance measures
# linear
# loglog025 eq. to (log(log(4*x)))
# log10
# sqrtlog10 eq. to sqrt(log_10(x))
#
# For a constant scale like M_Z use e.g.
# Mu1_Kernel OneNode
# Mu1_DistanceMeasure "log10"
# Mu1_NNodes 1
# -------------------------------------------------------------------- #
#X_Kernel Lagrange # (def.= pp: Lagrange; DIS: Catmull)
#X_DistanceMeasure sqrtlog10 # (def.= pp: sqrtlog10; DIS: log10)
#X_NNodes 20 # (def.= pp: 20; DIS: 18)
#X_NNodeCounting NodesPerBin # (def.=NodesPerBin)
#Mu1_Kernel Lagrange # (def.=Lagrange)
#Mu1_DistanceMeasure "loglog025" # (def.=loglog025)
#Mu1_NNodes 6 # (def.=6)
# Scale2 not used for fixed-scale tables
Mu2_Kernel OneNode # (def.=Lagrange)
Mu2_DistanceMeasure "loglog025" # (def.=loglog025)
Mu2_NNodes 1 # (def.=6)
0% Loading or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment