|chil home > projects > ACT-R/PM > documentation >|
A description of all the system parameters in ACT-R/PM. Paramater values can be viewed with
PM-SHOW-PARAMS and set with
:auto-dequeue t | nil
NIL. When a
!pop! occurs that clears the goal stack, but there are still events in the schedule queue, you may want those events to be dequeued. By default, ACT-R/PM will prompt with a question about whether or not the events should be dequeued. The prompting can be turned off and the events automatically dequeued by setting this value to
:cursor-noise t | nil
NIL. When this is NIL, cursor moves always land on exactly the specfied point (or object center). When this is T, there is some noise in the output position. The amount of noise depends on the width of the target object, but it will now miss most objects about 4% of the time.
Default 1.0. When a cursor movement is specified, to accurately compute the movement execution time, the system needs to know the size of the target in degrees of visual angle. RPM will compute this for object if it can, but when it can't, it will use this default.
NIL. The function supplied here will be called with the current simulated time as an argument every time RPM runs the ACT Cognition Layer. You can use this parameter to set up a function for dynamic displays that need to updated periodically, but it is recommended that you define a
device-update method for your device instead.
Default 0.300. Sets the content delay (in seconds) for auditorily-presented digits.
Default 0.600. Sets the duration (in seconds) for auditory presentation of a digit.
Default 0.500. Sets the recoding delay (in seconds) for auditorily-presented digits.
:incremental-mouse-moves t | nil
Default NIL. When set to T, this will update the mouse location every 50 ms or so when it is in motion. When set to NIL, it will update the mouse location only at the end of each mouse move.
0.100. It is possible for a Fitts-law movement to take a very short time. This sets the minimum movement time.
Default 0.05. This is the minimum time (in seconds) required for movement execution. (This parameter isn't new, but the default is now lower.)
Default 0.050. This is the time (in seconds) required for feature preparation for each feature.
Default 0.050. This is the time (in seconds) required for the Motor Module to initiate a movement.
Default 0.1. Determines the base coefficient used for Fitts' computations with mouse movements. The basic equation is time = a * log2( distance / width + 0.5), and this controls the a parameter.
:needs-mouse t | nil
Default T. Determines whether or not RPM should control the cursor location.
:optimize-visual t | nil
T. If set to
T, the Vision Module does not use rational feature-recognition but rather builds the visual icon entirely at the word level, assuming perfect recognition. If
NIL, then the Vision Module does it "the hard way." More detail can be found in the Principles subsection. Replaces
*optimize-visual* from the old Visual Interface.
:output-speech t | nil
NIL. If set to t, then RPM will produce synthesized audio (if the facilities are present) as a result of
0.075. Determines the base coefficient used for Fitts' computations with PECK-style movements. The basic equation is time = a * log2( distance / width + 0.5), and this controls the a parameter.
Default 72.0. For accurate computations involving visual angles, this parameter needs to reflect the number of pixels (dots) per inch of the screen. The default is the MacOS default, and may not be correct for your monitor. (Not really new for beta 6, but has a new name.)
:print-viewed t | nil
NIL. When this parameter is set to
T, then calls to
MOVE-ATTENTION will produce output detailing which chunks have been added to declarative memory.
:process-cursor t | nil
NIL. In many experiments, the mouse cursor is not used and
therefore should be ignored by RPM. If the cursor is going to be looked for
by Vision or moved by Motor, then this should be set to
:randomize-time nil or number
NIL. Normally, completion times for perceptual-motor operations have fixed times. To make those times vary randomly around those fixed values, set this to a positive integer. A value of 2 will produce an
output from 1/2 to 3/2 of the input for each time; a value of 4 will produce an output from 3/4 to 5/4, and so forth. If no number is supplied (e.g., this is just set to
T, then the value 3 is used, which is the EPIC default.
:real-time t | nil
NIL. Setting this switch to
T will cause ACT-R/PM to pause periodically so that it runs in something close to real time if it is running faster than real time. This was included for running demos.
NIL. The provided function will be called when RPM is running in "real time" mode and the system is waiting for real time to catch up with simulated time. The function is called with one argument, which is the disparity (in seconds) between ACT-R's clock and the wall clock.
Default 3.000. The amount of time after a sound has finished it takes for the sound to be deleted from the audicon.
Default 10.0. When a new sound chunk is created, it has some default base-level activation. This parameter controls that activation.
NIL. This function will be called whenever speech is output. It is recommended that you define a
device-handle-speech method for your device instead.
Default 0.150. Specifies the amount of time (in seconds) it takes to speak an "average" syllable. Used for computing the execution time for the speak command.
Default 0.050. Content delay for tones.
Default 0.285. Recoding time for tones.
:trace-modules t | nil
T. When this parameter is set to T, each Module will print a message whenever it runs a command. Useful for debugging and getting an idea when events are happening.
Default 15.0. For accurate computations of visual angles, the system needs to know how far the subject is from the screen. If the subject is closer or further than 15 inches, change this parameter. (Not new for beta six, but it has a new name.)
Default 0.185. This parameter controls how long it takes to shift attention in the visual field.
Default 0.5. When
MOVE-ATTENTION is called, it takes time between when the move is initiated and when it completes. If there is nothing at the specified location when this completes,
MOVE-ATTENTION will not find any object, even if the move was only one pixel. This parameter can be used to solve that problem, as this parameter specifies the distance (in degrees of visual angle) an object can move and still be seen by
Default 1.0. When
MOVE-ATTENTION is called, the attended object becomes an activation source until the next call to
MOVE-ATTENTION. This parameter determines how much activation that source has.
|People Facilities Research Projects Miscellany Search | Home|
Last modified 2004.03.12