Configuration Files

Text-based configuration files are used to feed parameters into swarm utility.

Here is a sample of configuration file:

# I/O configuration
input=sample.in
output=sample.out

# Integrator config
integrator=rkck_adaptive
min time step=0.0001
max time step=0.001
error tolerance=1E-27

Syntax is:

• Every line starting with # is ignored
• Empty lines are ignored
• A valid line is in form name=value
• Spaces in the parameter name and parameter value are allowed.
• If a name is assigned more than once, only the last occurence is considered.
• Parameter names are case-sensitive.

Module	Name	Default Value	Description
Swarm Executable (Utils)	nsys		Number of system when auto-generating the ensemble
	nbod		Number of bodies when auto-generating the ensemble
	spacing_factor	1.4	Number of bodies when auto-generating the ensemble
Swarm Executable	destination_time	10 pi	Marker for end of integration, all systems integrated and synchronized to this time
	interval	disabled	Intervals testing of stability of the system
	logarithmic	disabled	Logarithmic base for exponentially growing intervals (used for stability graphs in logarithmic scale)
	allowed_deltaE	0.01	Maximum deviation error from energy preservation allowed. If the maixmum energy conservation error is bigger than the number specified then the integration is stopped and marked as failed (Primarily used for automated stability test)
	pos_threshold	1e-10	Maximum deviation allowed for position magnitude when comparing an ensemble against the reference
	vel_threshold	1e-10	Maximum deviation allowed for velocity magnitude when comparing an ensemble against the reference
	time_threshold	1e-4	Maximum deviation allowed for system time when comparing an ensemble against the reference
	input		Binary input file
	output		Binary output file
	text_input		Text input file
	text_output		Text output file
SwarmNG library	nogpu	0	If set to 1, the GPU is not initialized. nogpu = 1 should be used when running Swarm without an actual GPU on the system.
	system_per_block	SHMEM_CHUNK_SIZE	Number of systems in a CUDA block
	CUDA_DEVICE	0	Number of the CUDA devices to use (Only used if there are more than one GPUs on a system)
Integrator	integrator		Name of The integrator plugin used for integration
	max_iterations		Maximum number of iterations in the integration kernel internal loop
	max_attempts		Maximum number of attempts on running the integration kernel to finish the integration
Fixed time step integrators	time_step	0.0001
Adaptive step integrators	min_time_step		Smallest step size allowed in adaptive mode
	max_time_step		largest step size allowed in adaptive mode
Adaptive step Hermite Integrator	time_step_factor		A factor that determines the effect of energy estimate on the time step
Adaptive step Runge-Kutta integrator	error_tolerance		Amount of error allowed for adaptive integration
Logging Subsystem	log_writer	null	Output method used for logging: null is to discard output bdb writes to Berkeley DB databes (recommended) binary writes binary files.
	log_output		For binary logger: path to the output file where the log is stored
	log_output_db		For bdb logger: path to the database file where the log is stored
Log interval monitor	log_interval		The fixed interval time at which the system is logged (if enabled)
Stop-on-Ejection monitor	rmax	+Infinity	Maximum allowed distance between a planet and the sun before the planet is marked as ejected
Stop-On-Collision monitor	collision_radius	0	The closest that two planets can get without triggerring a collision
Stop-On-Any Large distance monitor	stop_on_rmax		Should the monitor stop integration if there is a large distance
Stop-On-Close-Approach monitor	close_approach	0