Learn Git and GitHub without any code!

Using the Hello World guide, you’ll start a branch, write comments, and open a pull request.

Read the guide

Watch

Notifications

2
0

0
Fork

Fork sixdesk

If this dialog fails to load, you can visit the fork page directly.

0

apc-llc/sixdesk

Find file

Copy path

Find file

Copy path

dmikushin Initial commit of SixDesk r31 we have started working with some time … d2f0a38

Apr 30, 2014

1 contributor

Users who have contributed to this file

311 lines (271 sloc) 8.25 KB

Raw Blame History

	#!/bin/sh
	# Januray, 2012
	# NOTA BENE: there is a new option for a more precise proton mass
	# and also a new flag to ignore tunes (tune=0)
	# and a new bunch_charge option for both collision and injection

	export pmass=938.271998
	# or you can use the new more precise value
	# by uncommenting the next line
	#export pmass=938.272013

	# New bunch_charge variable for fort.3.mother1_[col/inj]
	export bunch_charge=1.1000e+11

	# The name of the "study" .le. 16 characters ([a-z A-Z 0-9 + - _ . ,]
	# It must NOT contain a "%" (replacing "#") nor "_""_"
	# It now includes the Version info e.g. v64lhc
	# (If running mad6t you must have a $LHCDescrip.mask file in the subdirectory mask)
	export LHCDescrip=lhc

	# The workspace for the scripts, sixdesk.log and links to sixtrack_input, track and work,
	# It must NOT contain a "%" (replacing "#") nor "__" (two consecutive underscores)
	export workspace=w1

	# The directory for the workspace (normally backed up) e.g. $HOME
	initial=`echo $LOGNAME \| cut -c 1`
	export basedir=/afs/cern.ch/user/$initial/$LOGNAME

	# The scratch directory for temporary input files, jobs, sixdesk databases etc (not normally backed up)
	export scratchdir=/afs/cern.ch/user/$initial/$LOGNAME/scratch0

	# The root track directory for all the tracking input and output
	# This directory can be used by multiple workspaces for scalability
	# but the default now includes the workspace name (track is added
	# by sysenv to make $sixdesktrackdir and the LHCDescrip is added
	# as necessary by run_six etc etc).
	export trackdir=$scratchdir/$workspace

	# The backed up or scratch directory for the mad6t input
	# It may be set to an existing sixtrack_input if you do not need to run mad6t.
	export sixtrack_input=$scratchdir/sixtrack_input/$workspace/$LHCDescrip

	# Flag for run_six for re-running cases
	# 0 => Jobs will not be re-submitted,
	# 1 => Jobs will not be re-submitted if fort.10.gz exists
	# 2 => Jobs will be re-submitted anyway using existing input files
	# (3 => Input files will be deleted, re-generated and jobs re-submitted anyway)
	export sixdeskforce=0

	# One and one only of the following options must be selected
	# BOINC for LHC@home, LSF for LSF batch, CPSS for Windows desktops at CERN, or GRID
	export platform=LSF

	# The message logging level: 0 => Start/Stop summary only, 1 => + error messages,
	# 2 => + informatory messages, 4 => + debug information (additive)
	export sixdesklevel=3

	# sixdeskecho = "" => only minimum information is echoed to stdout
	export sixdeskecho="yes please"

	# Options for run_mad6t (but sixtrack_inputdir defined above)
	# If CORR_TEST = 1, check_mad6t will copy all the corrector strengths
	# for each seed into one file in sixtrack_inputdir
	export CORR_TEST=0

	#If fort_34 is null, the fort.34 files will not be copied to sixtrack_inputdir
	export fort_34=""

	# Start and end seed for mad6t
	export istamad=1
	export iendmad=60

	# Note that all mad6t runs are with the local LSF batch system,
	# even if tracking is being performed on BOINC, CPSS or the GRID,
	# BUT they can now be done interactively with run_mad6t -i !!!
	# LSF batch queue for mad6t jobs. Normally 8nm is enough, but for
	# lhc.D1-D2-MQonly-inj-no-skew study 1nh was necessary.
	export madlsfq=1nh
	export shortlsfq=8nh
	export longlsfq=1nd
	export dalsfq=1nd

	# Format: "col" or "inj" ( also for run_six)
	export runtype=inj
	# and the beam can be set to "" or b1 or B1 for beam1, or to b2 or B2 for beam2
	export beam=""

	# Options for run_six
	# idfor = 0 => closed orbit added, = 1 => initial co-ordinates unchanged
	export idfor=0

	#The algorithm for beam-beam interactions
	# 0 is the "old" way, "1" is the new Erskine/McIntosh algorithm
	export ibtype=0

	#general parameters
	# No of particle pairs, normally 30 (other values ARE NOW SUPPORTED
	# in the range 1 to 32
	export sixdeskpairs=30

	# seeds for everything except mad6t
	export ista=$istamad
	export iend=$iendmad

	# angle step
	export kstep=1

	#normalized emittance
	export emit=3.75

	# e0 gamma
	if test $runtype = "inj"
	then
	export e0=450000.
	export gamma=479.6
	elif test $runtype = "col"
	then
	export e0=7000000.
	export gamma=7460.5
	fi

	#phase space dimension
	export dimen=6

	#switch for chromaticity correction
	# If chrom=0 no correction is applied but two one turn runs are made
	# using chrom_eps to generate sensible values (for info only).
	# If chrom=1 then chromx and chromy are used
	export chrom=0
	export chromx=2.
	export chromy=2.
	export chrom_eps=0.0000001

	#initial relative momentum deviation
	if test $runtype = "inj"
	then
	export dpini=0.00075
	elif test $runtype = "col"
	then
	export dpini=0.00027
	fi
	export dpmax=0.002

	#tunes
	# If tune=0, the tunes are ignored.
	export tune=1

	# Multiple tunescans can be specified
	# tunescans will be computed on a straight line from
	# (tunex,tuney) with gradient deltay/deltax until and including (tunex1,tuney1)
	# The tunes must be 10 <= tune < 100 in format dd.dd[d][d]
	if test $runtype = "inj"
	then
	#tunes (injection)
	export tunex=64.28
	export tuney=59.31
	export deltax=0.001
	export deltay=0.001
	export tunex1=64.28
	export tuney1=59.31
	elif test $runtype = "col"
	then
	#tunes (collision)
	export tunex=64.31
	export tuney=59.32
	export deltax=0.001
	export deltay=0.001
	export tunex1=64.31
	export tuney1=59.32
	fi

	#switch for Guignard resonance calculation
	export reson=0

	#switch for precise tune determination
	export sussix=0
	# Note if sussix = 1, short must be 1 (long/da 0)
	# and kini = 1 FRS??????

	# Now we have short or long term parameters
	# only one of short and long may be 1
	#short term parameters
	export short=0
	if test $short -eq 1
	then
	export ns1s=1
	export ns2s=10
	export nss=10
	#
	export turnss=1000
	export turnsse=3
	export writebins=1
	# kini=0 is the momemtum detuning and then set to 1
	export kini=0
	export kend=5
	export kmax=5
	# and set up the LSF job type and queue
	export lsfq=$shortlsfq
	export lsfjobtype=sixmedium
	fi

	#long term parameters
	export long=1
	if test $long -eq 1
	then
	# amplitudes which must be 0 <= amplitude < 100 in format [d][d][.][d][d][d]
	export ns1l=4
	export ns2l=12
	export nsincl=2
	#
	export turnsl=100000
	export turnsle=5
	export writebinl=1000
	export kinil=1
	export kendl=5
	export kmaxl=5
	# and set up the LSF job type and queue
	export lsfq=$longlsfq
	export lsfjobtype=sixtracking
	fi

	#da parameters

	export da=0
	if test $da -eq 1
	then
	export sixdeskM="-M 1000000"
	export lsfq=$dalsfq
	export lsfjobtype=sixda
	# some physics needed here
	# dimda = 4 or 6 but MUST be set to NOT null for short/long
	export dimda=4
	export NO=4
	((NO1=NO+1))
	export NO1
	export NV=5
	export turnss=1
	else
	export dimda=0
	fi

	# Parameters for run_join10
	# Range of tracked turns
	if test $short = 1
	then
	export turnsemax=$turnsse
	fi
	if test $long = 1
	then
	export turnsemax=$turnsle
	fi

	# For run_post
	if test $long = 1
	then
	export turnse=$turnsle
	fi
	if test $short = 1
	then
	export turnse=$turnsse
	fi

	# flag for short and long plotting
	# iplot=1 produces plots for each seed
	# iplot=2 prints as well using $PRINT_CMD
	export iplot=0

	# kvar=0 => kend=$kendl, kvar=1 => kend=$kinil
	# kvar should be set to 1 to obtain the DA
	# as a function of the angles (when long=1)
	# The DA over ALL seeds and angles is plotted for each angle
	# even if iplot=0
	if test "$long" -eq 1
	then
	export kvar=1
	else
	export kvar=0
	fi
	# BNL variables for bnlelens
	# Later I could separate out elens and bnl dependencies in sixtrack.s
	# as we may wish to run LHC elens studies
	export BNL=""
	if test "$BNL" != ""
	then
	# We expect to find beambeamdist.dat, fort.2, fort.3.mother and fort.8 and fort.16
	# in sixtrack_input and run_bnl will check this
	#
	# The "seed" 1000 for example as used by Yun Luo
	export bnlseeds=1000
	export bnlseede=1000
	# First and Last sample number in beambeamdist.dat
	export bnlsamples=1
	export bnlsamplee=1
	# Number of pairs descriped in a sample in beambeamdist.dat
	export bnlpairs=32
	# Number of pairs for each job
	# This used to generate bnlpair names like 1-2, 2-3, and so on....63,64
	export bnlperjob=2
	# Now the turns, a bit tricky, since we use nwri for both deciding how often
	# to checkpoint, and how often to write beambeam-output.dat!
	# I am assuming we run 2 pairs per job and then the values below
	# are reasonable
	export bnlturns=10000000
	export bnlturnse=e7
	export bnlwrite=100000
	# for my testing I used
	export bnlturns=50000
	# end BNL variables
	fi
	#

Edit status

Suggestions:

Learn Git and GitHub without any code!

Fork sixdesk

apc-llc/sixdesk

Users who have contributed to this file

Learn Git and GitHub without any code!

Fork sixdesk

apc-llc/sixdesk

sixdesk/sixjobs/sixdeskenv.bnl

Users who have contributed to this file