usage: pyXplor [option] [file]
where option is one of:
-c <command> - invoke the Python interpreter, running
the specified command.
-m module... - invoke the Python interpreter, running
the specified module.
-help - print this message.
-version - print version info
-features - print version info, and a list of features
If no file is specified on the command-line, commands will be read from stdin.
Common options:
-debug <cmd> - run xplor under debugger given by cmd
-notrace - turn command echoing off (set echo=off)
-quiet - turn off verbose output (set messages=NONE)
-nice <#> - specify a nice (see man nice) increment
-omp <#> - specify the number of OpenMP threads to use.
This overrides the value of the environment
variable OMP_NUM_THREADS.
The default value is 1.
-fpe - enable detection of floating point exceptions.
If an overflow, divide by zero occurs, the
program will halt and core dumped
Options for VMD-XPLOR:
-host - specify host on which VMD is running -port - specify port for connection to VMD
Options for parallel jobs:
-parallel - spawn multiple jobs on remote hosts
-machines <spec> - The spec argument is a space-separated
list of machines on which to run, or the
name of a file with one machine name per
line (lines beginning with the # character
are ignored as comments). The name
localhost'' is special in that it causes
a new processes to be forked on the local
machine. Each instance of a machine name
causes one process to be run, and if the
names are specified consecutively, one rsh
connection is made, and a process forked for
the number of times the machine is
specified. This option overrides the
MACHINES environment variable.
-crash_on_startup - specify this flag if you wish a parallel
calculation to fail at startup if any
process fails to startup. The default
behavior is to continue the full calculation
using only those processes which start.
-startup_delay <#> - specify a delay factor for processes
connecting to the server (process 0).
The delay time is the factor multiplied by
the process number.
-pport <port #> - port for parallel communication. This
overrides the environment variable
XPLOR_PPORT. By default this value is
9021 + PID mod 1000, where
PID is the process id.
-rsh <cmd> - command to connect to remote machine.
This overrides env var XPLOR_RSH
[default value: ssh -n -x]
-slurm - used when called by slurmXplor. Specifies
that srun is used to launch each process
individually instead of forking N local
processes on each node.
-scyld <#> - on a Scyld cluster, specify the number of
structures to run in parallel. This option
overrides the environment variable
SCYLD_PROCESSES.
-smp <#> - on a multiprocessor node, specify the number of
structures to run in parallel. This option
overrides the environment variable
SMP_PROCESSES.
-o <file> - specify prefix for log file.
name will be file.XPLOR_PROCESS
Options for parallelization of ensemble calculations:
-num_threads <#> - number of processors to use in
ensemble calculations. This option
overrides the environment variable
NUM_THREADS
-thread_filename <prefix> - prefix for thread output (used to
debug for ensemble calculations).
Each non-zero ensemble member will
have an output file named
prefix-schwitrs-UID-i, where i is the
ensemble member index.
usage: pyXplor [option] [file]
where option is one of:
-c <command> - invoke the Python interpreter, running
the specified command.
-m module... - invoke the Python interpreter, running
the specified module.
-help - print this message.
-version - print version info
-features - print version info, and a list of features
If no file is specified on the command-line, commands will be read from stdin.
Common options:
-debug <cmd> - run xplor under debugger given by cmd
-notrace - turn command echoing off (set echo=off)
-quiet - turn off verbose output (set messages=NONE)
-nice <#> - specify a nice (see man nice) increment
-omp <#> - specify the number of OpenMP threads to use.
This overrides the value of the environment
variable OMP_NUM_THREADS.
The default value is 1.
-fpe - enable detection of floating point exceptions.
If an overflow, divide by zero occurs, the
program will halt and core dumped
Options for VMD-XPLOR:
-host - specify host on which VMD is running -port - specify port for connection to VMD
Options for parallel jobs:
-parallel - spawn multiple jobs on remote hosts
-machines <spec> - The spec argument is a space-separated
list of machines on which to run, or the
name of a file with one machine name per
line (lines beginning with the # character
are ignored as comments). The name
localhost'' is special in that it causes
a new processes to be forked on the local
machine. Each instance of a machine name
causes one process to be run, and if the
names are specified consecutively, one rsh
connection is made, and a process forked for
the number of times the machine is
specified. This option overrides the
MACHINES environment variable.
-crash_on_startup - specify this flag if you wish a parallel
calculation to fail at startup if any
process fails to startup. The default
behavior is to continue the full calculation
using only those processes which start.
-startup_delay <#> - specify a delay factor for processes
connecting to the server (process 0).
The delay time is the factor multiplied by
the process number.
-pport <port #> - port for parallel communication. This
overrides the environment variable
XPLOR_PPORT. By default this value is
9021 + PID mod 1000, where
PID is the process id.
-rsh <cmd> - command to connect to remote machine.
This overrides env var XPLOR_RSH
[default value: ssh -n -x]
-slurm - used when called by slurmXplor. Specifies
that srun is used to launch each process
individually instead of forking N local
processes on each node.
-scyld <#> - on a Scyld cluster, specify the number of
structures to run in parallel. This option
overrides the environment variable
SCYLD_PROCESSES.
-smp <#> - on a multiprocessor node, specify the number of
structures to run in parallel. This option
overrides the environment variable
SMP_PROCESSES.
-o <file> - specify prefix for log file.
name will be file.XPLOR_PROCESS
Options for parallelization of ensemble calculations:
-num_threads <#> - number of processors to use in
ensemble calculations. This option
overrides the environment variable
NUM_THREADS
-thread_filename <prefix> - prefix for thread output (used to
debug for ensemble calculations).
Each non-zero ensemble member will
have an output file named
prefix-schwitrs-UID-i, where i is the
ensemble member index.