Starting and Restarting

KEPLER is initiated by typing:

KEPLER NAMEP ( GENERATOR | DUMP ) ( OPTION )*

kepler s25s2a s25s2ag s

KEPLER

is the executable which is to be run.

The version used for a given run and its date are recorded at the beginning of each ASCII output file.

NAMEP

is the character name given to the current problem. This name is also used to begin the names for various output and dump files created during this problem (see below).

GENERATOR | DUMP

is a character name of either a generator file (in which case its name must end in g (e.g., s25s2ag), or a restart dump (in which case its name must not end in g (e.g., s25s2az or s25s2a#10350 or s25s2a#presn).

Special Cases	Meaning
g	To generate from file NAMEPg.
z	To restart from file NAMEPz,
p	To restart from file NAMEP(with last character deleted)#presn.
f	To restart from file NAMEP(with last character deleted)#final.
z?	similar to z, z1? z9 will star up dumps with that ending. The creation of these files is controlled by nzsave (p 488).
#*	start from labeled dump with same base name.
*g, *z, *#*	separate out base name for run automatically.

Changed In Version <15: Added z[0-9]* and # options and automatic base name separation.

OPTION

(optional) is a set of one or more space delimited words in arbitrary order chosen from among the following possibilities:

Option	Meaning
s	To suspend the code after starting.
k	do not load BURN data and kill burning, similar to the killburn command.
d	To destroy the ASCII output files after sending them to the microfiche printer (not currently implemented – see also the discussion of parameter iautoout (p 159).
h	To automatically make a paper copy of the ASCII output files (not currently implemented).
c	write out stored command file (overwrite existing).
a	append stored command file.
p	make plot on start-up.
n	suppress graphical output, e.g., for batch mode.
x	DEBUG mode. Allow running without proper git hash.
X	Extended DEBUG mode. Could be used to enable internal debugging code. Adds 2 to internal idebug variable.
e	Allow experts to overwrite of ittyv (p 538) by command files. This is for the % ittyv command file directive.

Changed In Version >15: Added k option.

Changed In Version 17.1: Added c and a options.

Changed In Version 17.6: Added x option.

Changed In Version 17.11.7: Added p and n options.

Changed In Version 18.0.1: Added X option.

Changed In Version 18.37.1: Added e option.

Warning

Due to automatic name decomposition, run names must not end in the characters g or z.

Initial Operation

In the absence of any OPTIONs, the code will begin calculations imniediately and keep its ASCII output files on disk. It will either generate from cycle 0 if NAMEGIZ is the name of a valid generator file or restart a previous problem if NAMEGIZ is the name of a restart dump compatible with this version of KEPLER. The code will attempt to restart from restart dumps created by older versions of KEPLER by setting any newly added parameters to their default values, and printing warning messages concerning possible incompatibilities. Such compatibility, however, is not guaranteed, and it is the user’s responsibility to carefully check whether such a restart operation has been perfonned correctly (see subroutine restart). If equation of state routine changes have been made in the interim, the newe or eostrans commands may be helpful.

Required Input Files

The current working directory must contain the specified compiled version of KEPLER, a link to a current version of KEPLER or KEPLER is in the PATH specification. The named generator or restart file must also be physically resident in the current directory.

All other files KEPLER needs should either be in the local dicrectory (searched first), in the path set in the datapath variable using the datapath card in the generator (searched second), or in the path provided in the KEPLER_DATA environment variable. In bash, e.g., in the .bashrc file, this can be set using the command

export KEPLER_DATA=${HOME}/kepler/local_data/

Note

The trailing / in the path spcification needs to be supplied.

Specically, if BURN is used the generator file specified in the generator file needs to be found.

If BURN co-processing has been specified, the directory in which KEPLER is in must also contain the BDAT data file, bdat, which contains the needed nuclear cross-sections for detailed isotopic networks. This file by default is in $HOME/kepler/source/data or $HOME/kepler/local_data and can be soft-linked to the current working directory by the command

ln -s ~/kepler/local_data/bdat .

Further files that are usually needed include

solabu.dat
branch.dat
nuccedat
nuncdat
nuccecdat
adapnet.cfg
opal95.dat
alamos1.dat
alamos2.dat
alamos3.dat
alamos4.dat
wdat
decay.dat
condall06.dat

In order for MONGO plots to work, KEPLER needs to know where to locate the corresponding font files. The easiest way is to set environment variables to their location. Usually, these files should be located in the subdirectory ~kepler/mongo. In bash these can be set using

export HELPFILE=$HOME/kepler/mongo/help.dat
export MONGOPS=$HOME/kepler/mongo/postscript/
export FONTDAT=$HOME/kepler/mongo/fonts.dat
export FONTNEW=$HOME/kepler/mongo/fonts.vis

Note

All see settings can also be done by the initialisation script, which should conveniently be called from the .bashrc file:

source ${HOME}/kepler/source/init.sh

See also Setup.

KEPLER writes all I/O in big endian format. For the Intel compiler one should therefore set the environment variable F_UFMTENDIAN accordingly. In bash, this can be done using the command

export F_UFMTENDIAN big

Debug and Production Runs

By default KEPLER will no longer run if built from a non-clean git status of the source directory. You can for running force debug purpose if you specify the x flag on the command line.

Important

The purpose of this change is to ensure code git hashes for the KEPLER version running a problem remain accurate. Results from non-clean git repos can never be traced back later with any level of confidence.

Added In Version 17.6.

qlib Files

Deprecated Since Version >15.

The qlib file and the qq post-processor dump file corresponding to the named restart dump (see below for definitions) are also required for restarting a problem making post-processor dumps. Note that KEPLER will overwrite information generated by the previous problem for duplicated cycles. Post-processor dump files not containing the restart cycle may also be in the working directory, but are not required.

Postprocessor Files

Deprecated Since Version >15.

If post-processor dumps have been requested in the generator file (or less desirably initiated by a command from the terminal), KEPLER will create a (qlib) library file with a name in the form NAMEP.lib (e.g., :code`s25s2a.lib`) whith maintains a record of zonal mass coordinates and stores a time history of the time edit variables. A series of post-processor dump (qq) files are also created starting with a file named in the form NAMEP.qa (e.g., s25s2a.qa) which stores a compressed time history of the dump variables specified by dump commands. These files are updated every ncycqq (p 299) cycles. The qlib library file can store information for up to 20000 * ncycqq (p 299) KEPLER cycles and up to 17999 distinct grid points before becoming full. For ease in handling, however, the post-processor dump files are limited to a size of lenqmax (p 44) whose default value is 3000000 bytes. New post-processor dumps will be created as required with suffixes in the sequence: .qa, .qb, .qz, .q0, … .q9, .ra, .rb, .rz, .r0, … .r9, etc. The qlib library file must always stay on-line while the problem it corresponds to is running, but all but the most recent post-processor dump (qq) files may be stored elsewhere if disk space is a problem. This will limit the range of possible look and timemap commands, however.

Restart Dump Files

KEPLER will create a running restart (z) file named in the form NAMEPz (e.g., s25s2az), which it overwrites every ndump (p 18) cycles. It will also create labeled restart dumps with names in the form NAMEP#ncyc (e.g., s25s2aZ#12750) every ndump (p 18) * nsdump (p 156) cycles, where where ncyc is the current cycle number. If a file with the name NAMEP#ncyc already exists, it will be overwritten.

Ascii Output Files

KEPLER will also create ASCII output files with names in the form NAMEP_NCYCO (e.g., s25s2az_O), where NCYCO is the cycle number at which the problem is being generated or restarted. To prevent such files from becoming too large, however, the current output file is closed whenever mod(NCYC ,nnewoutf (p 197)) = 0 (by current default every 2000 cycles) and a new output file with a name in the form NAMEP_NCYC (e.g., s25s2az_2000) is created. These types of ASCII output file will overwrite any file by the same name that might already exist.

Dump-Grid Status Files

Deprecated Since Version <15.

If post-processor dumps are being made, KEPLER will also alternately create, read, and destroy files with names in the form NAMEP.0 and NAMEP.1 (e.g., s25s2a.0 and s25s2a.1) which contain information about the grids currently being used to represent each requested dump variable. This reduces the size of the KEPLER executable file at the cost of some extra files on disk. If for any reason (restart, disk crash, etc.), these flies are not available, KEPLER will automatically generate a new set of dump grids along with a warning message. This is not really a problem except that it takes up a little extra space in the post-processor dumps.

Running in Background

On UNIX systems input and output can be redirected. On some systems, it may also be needed to use nohub. Or you can run in a virtual screen, such as provided by screen. No graphical output should be use for this, obviously, as it will crash the job when disconnected.

nohup KEPLER NAMEP ( GENERATOR | DUMP ) ( OPTION )* </dev/null 2>&1 > OUTFILE

Here OUTFILE may also be /dev/null if you do not care about the output.

Note

Do not use “stop” option s.

You may continue to look at output using

tail -f OUTFILE