Input to a RatNetwork comes from a variety of sources: vision, sound, proximity to markers, proprioreception and place, each of which appears at a RatNetInputNode (place input is a short cut to generate place cells directly). The input nodes can optionally rescale the raw input onto a range vMin to vMax. Without rescaling, input is normalized to the range [0, 1]. Rescaling can be useful, for example, to map input activity to a membrane potential for continuous input to an integrate and fire population.
There is one special input node, named posn pr position which carries position information. This is not intended for use in the network (although there is no structural reason reason why it should not be), but rather as a way of recording the rat's position along with its network activity.
The datea arriving at this node comes from the projection of the rats position onto a specified line in the environment which is selected by setting the projectOnto flag in the desired line - see RatEnvironmentLine. It can be used in the spike analysis component to correlate spiking activity with progress along a path through the environment.