Sandhya, probably the easiest approach would be along these lines:

1. Create a Continuous2D for your agents that are moving from node to  
node.
2. Each timestep, each of these agents do the following:
	a. Look up the 2D locations of the two nodes (A and B) on either end  
of the edge
	b. Compute how "far" (from 0.0 to 1.0) the agent should presently be  
from A to B.  Call that f.
	c. Set the agent's location to be
		< f*pos_x(A) + (1-f)*pos_x(B), f*pos_y(A) + (1-f)*pos_y(B) >
3. Create a Continuous2D portrayal for your continuous2D of these  
agents, and put it in the front of the other portrayals.  You're in  
business.

On Mar 29, 2008, at 10:21 PM, Sandhya Arunachalam wrote:

> Hi all
>
> I have attached with this post an image of my network . Can anyone  
> please
> help me in creating an agent that moves from node to node through  
> the edge
> connecting them ?
>
> Even the smallest suggestion as to where to start would be helpful.
>
> Thanks
>
> Sandhya