|COGENT Version 2.2 Help|
When a model is opened via the Research Programme Manager COGENT displays the model in its own window. Figure 1 shows an example.
Any box on the diagram may also be opened by double clicking on it (or right-clicking and selecting Open from the menu that pops up). Again, this will result in a new window appearing, with the new window containing the details of the box that has just been opened.
All COGENT windows that display box contents have the same basic layout. They consist of three horizontal sections (recall Figure 1):
The buttons on the left of the menu bar have the following functions:
|File|| A menu providing access to save, print
and syntax checking facilities. Save may be applied to save the
contents of a single box or, if the box is a compound, it may be applied
recursively to also save the contents of all boxes contained within the
Print...opens a dialogue for printing selected parts of the box -- see the separate documentation. Check Syntax examines the contents of the box for common errors -- see the separate documentation.
|Edit||A menu providing access to editing commands (find, replace, undo and redo), a check consistency command (to check that the box contents are consistent with arrows to/from that box) and the preferences panel.|
|Run||A menu providing access to commands for executing a model. These commands allow model initialisation, stepping through one cycle of the execution model, running the model until it terminates, halting the execution of the current model, and restarting the execution engine should it crash.|
|OOS...||Open the OOS window. This window provides access to the script which controls model execution, the script editor, and panels which display various kinds of information relating to model execution.|
|Help...||Open a web browser display a help file. If the help preferences are configured correctly, the browser will display the help file associated with the class of the current box.|
|Done||Close the current window. If any edits to the current box have not been saved, you will be prompted to save them before the window is closed.|
The icon short-cuts on the right of the menu bar have the following functions:
|Save all details of the current box. (Equivalent to File -> Save -> Save Box.)|
|Save all details of the current box and, if it is a compound, of all boxes contained within it. (Equivalent to File -> Save -> Recursive Save.)|
|Open the print dialogue to allow printing of specific aspects of the current box (and its sub-boxes). (Equivalent to File -> Print....)|
|Open the Find window to allow searching for terms within the current box (and any of its sub-boxes). (Equivalent to Edit -> Find.)|
|Open the Replace window to allow replacing of terms within the current box (and any of its sub-boxes). (Equivalent to Edit -> Replace.)|
|Undo the last edit made to the current box. (Equivalent to Edit -> Undo.)|
|If the last edit to the current box was made through Undo, then reverse that edit. (Equivalent to Edit -> Redo.)|
|Initialise execution of the current model. (Equivalent to Run -> Initialise.)|
|Run the current model for one step, using either the default script (which runs a single trial) or the script which has been specified for the model within the OOS window. (Equivalent to Run -> Step.)|
|Run the current model until it terminates, using either the default script (which runs a single trial) or the script which has been specified for the model within the OOS window. (Equivalent to Run -> Run.)|
|If the current model is running, send the process the stop signal. Execution will stop at when the current step is completed. (Equivalent to Run -> Stop.)|
Display of icon short-cuts may be disabled by setting the Show toolbars on box windows preference to FALSE.
Three pieces of information are displayed in the horizontal strip immediately beneath the menu bar: the box's name, description and type (or class). All three may be edited by the widgets in this region of the window. Note however that changing the type of a box may delete information that has previously been specified for that box if the information is not relevant to the new type.
The notebook occupies the largest portion of a box's window. It contains a series of pages which allow viewing or editing different kinds of information associated with the box. The precise pages available in the notebook depend on the type of box. The box shown in Figure 1 above, for example, is of type Compound/Generic. Correspondingly the notebook has pages labelled Diagram, Description, Properties, Message Matrix and Messages. Windows corresponding to boxes of type Buffer/Propositional, by contrast, have pages for Initial Contents, Description, Properties, Current Contents, Display Rules, Current Display and Messages. The various types of page and their respective functions are as follows:
Activation Graph: Displays a constantly updated graphical representation of the current activation values of nodes within an interactive activation network.
Current Contents: Displays a constantly updated text representation of the current contents of all types of buffers.
Current Display: Displays a constantly updated graphical representation of the current contents of propositional buffers, where the mapping between propositional buffer elements and graphical elements is given by the rules specified on the propositional buffer's Display Rules page.
Current Graph: Displays a constantly updated graphical representation of the current contents of graph buffers and graphical data sinks.
Current Image: Displays a constantly updated representation of the current contents of an analogue buffer.
Current Knowledge: Displays a constantly updated text representation of the current contents of a knowledge base.
Current Stack: Displays a constantly updated visual representation of the current contents of a stack buffer.
Current Table: Displays a constantly updated visual representation of the current contents of a table buffer or a tabular data sink.
Description: An area in which plan text describing the box may be entered and saved. This page is available for all types of box.
Diagram: An area for displaying and editing a box and arrow diagram. This page is available for all compound boxes (i.e., for all boxes that contain box and arrow diagrams), as well as for the top-level box in a model (which is itself a compound box).
Display Rules: An editable canvas of rules for translating propositional buffer elements into elements for display within an image (such as lines, boxes and circles). This page is only available on windows associated with propositional buffers.
Initial Contents: A canvas on which the initial elements of buffers (propositional, analogue, stack, tabular and graphical) and data sources may be entered. During model execution these elements are loaded whenever the box is initialised.
Initial Knowledge: A canvas on which the initial elements of a knowledge base may be entered.
Initial Nodes: A canvas on which the (initial) nodes and values of an interactive activation network may be entered.
Input Data: A canvas on which the initial data elements of a data source may be entered.
Message Matrix: An area on which a dynamically updated "matrix" showing messages passing between boxes is shown. The page is only available on notebooks corresponding to compound boxes, and the rows and columns of the matrix correspond to the boxes contained within the compound.
Messages: A dynamically updated display of messages received by the box during model execution. This page is available on windows corresponding to virtually all box types.
Properties: A page that displays the properties (and the values of those properties) associated with a box. Properties control configurable aspects of the appearance and behaviour of most types of box, and the properties page, which allows properties to be adjusted, is available on virtually all box windows.
Results: A dynamically updated text display of data received by a data sink. This page is available on windows corresponding to data sink boxes (i.e., text, tabular and graphical data sinks).
Rules & Condition Definitions: A canvas available on windows corresponding to process boxes on which rules and condition definitions, that define the behaviour of a process, may be displayed and edited.
Weight Matrix: A dynamically updated display of the current weight matrix of a feed forward network.
|COGENT Version 2.2 Help|