The Complete SPRLIB & ANNLIB

LAYER: - layer datatype

SYNOPSIS

typedef struct layer

{

char LayerName[80]; A description of the layer.

long LayerId; Layer identification number.

long LayerFlag; Layer flag - see LAYER-flags.

long NumUnits; Number of units in this layer.

long NumMaps; Number of unit-groups in layer.

MAP *FirstMap; Pointer to first group in layer.

MAP *LastMap; Pointer to last group in layer.

UNIT *FirstUnit; Pointer to first unit in layer.

UNIT *LastUnit; Pointer to last unit in layer.

union system_data SystemData; Union of data for algorithms.

union system_data UserData; Union of data for library user.

struct layer *Next; Pointer to next layer.

struct layer *Prev; Pointer to previous layer.

} LAYER;

DESCRIPTION

This type is used to represent a layer, currently only for shared weights neural networks. Most of the fields speak for themselves.

NOTE

All layers are stored in a double linked list. The LAYER structure is usually filled by a network creation function (create_sharednet).

The Complete SPRLIB & ANNLIB

`LAYER`

- layer datatype

SYNOPSIS

DESCRIPTION

NOTE

SEE ALSO

`char`	`LayerName[80];`	A description of the layer.
`long`	`LayerId;`	Layer identification number.
`long`	`LayerFlag;`	Layer flag - see `LAYER-flags`.
`long`	`NumUnits;`	Number of units in this layer.
`long`	`NumMaps;`	Number of unit-groups in layer.
`MAP`	*`FirstMap;`**	Pointer to first group in layer.
`MAP`	*`LastMap;`**	Pointer to last group in layer.
`UNIT`	*`FirstUnit;`**	Pointer to first unit in layer.
`UNIT`	*`LastUnit;`**	Pointer to last unit in layer.
`union system_data`	`SystemData;`	Union of data for algorithms.
`union system_data`	`UserData;`	Union of data for library user.
`struct layer`	*`Next;`**	Pointer to next layer.
`struct layer`	*`Prev;`**	Pointer to previous layer.