NAME

run - Run Soar

SYNOPSIS

run [n | 'forever'] [unit] [-self]

DESCRIPTION

This command runs the Soar agents. If the [-self] flag is issued, only the agent issuing the run command will run. n is a single integer which specifies the number of times to run Soar. If n is unspecified and no units are given, OR if n is the single keyword 'forever', then Soar is run until halted by problem-solving completion or an external interrupt. If n is specified, but no units are specified, then soar is run by decision cycles. If units are specified, but n is unspecified, then n defaults to '1'. The unit argument indicates the unit of measure to be used in counting Soar run cycles. unit must be one of the following keywords:

p
run Soar for n phases. A phase is either an input phase, preference phase, working memory phase, output phase, or decision phase.
e
run Soar for n elaboration cycles. (For purposes of this command, decision phase is counted as an elaboration cycle.)
d
run Soar for n decision cycles
s
run Soar until the nth time a state is selected
o
run Soar until the nth time an operator is selected
out
run Soar until the nth time output is generated by the agent
<s>
run Soar until the current level of subgoaling has terminated
<ss>
run Soar until the superstate's level of subgoaling has terminated
<sss>
run Soar until the supersuperstate's level of subgoaling has terminated
<o>
run Soar until the nth time an operator is selected at this level of subgoaling or until the current level of subgoaling is terminated.
<so>
run Soar until the nth time a superoperator is selected or until that level of subgoaling is terminated.
<sso>
run Soar until the nth time a supersuperoperator is selected or until that level of subgoaling is terminated.

Unlike the Soar 6 go command, run has no memory of settings from previous run commands.

There are two predefined aliases for the run command, provided as a convenience for use on the command line:

d [n]
run by decision cycles, equivalent to "run n d". n defaults to 1.
e [n]
run by elaboration cycles, equivalent to "run n e". n defaults to 1.

EXAMPLES

run
run all agents until halted by problem-solving completion or an external interrupt is given.
run 5 d
run all agents for 5 decision cycles
run d -self
run the current agent only for 1 decision cycle
run 3
run all agents for 3 decision cycles
run 2 e
run all agents for 2 elaboration cycles
run 1 s
run all agents until the next state is selected (i.e., until the next time an impasse arises)
run <so>
run until the next superoperator is selected (or until the superstate goes away)
run 3 <o>
run for 3 operator selections at this level (continuing through any subgoals that arise)

WARNING

In Soar8, the preference phase is no longer separated from the working memory phase, Soar continues right through from preferences into working memory without stopping. The phases are Input, Propose, Decide, Apply, Output; Propose and Apply each consist of at least one, but possibly many, preference/working-memory subphases.

SEE ALSO

d, e, stop-soar, predefined-aliases