Class for applying masks to layers. More...

#include <connection_creator.h>

Public Member Functions
	MaskedLayer (Layer< D > &layer, Selector filter, const MaskDatum &mask, bool include_global, bool allow_oversized)
	Regular constructor. More...

	MaskedLayer (Layer< D > &layer, Selector filter, const MaskDatum &mask, bool include_global, bool allow_oversized, Layer< D > &target)
	Constructor for applying "converse" mask to layer. More...

	~MaskedLayer ()

Ntree< D, index >::masked_iterator	begin (const Position< D > &anchor)
	Iterate over nodes inside mask. More...

Ntree< D, index >::masked_iterator	end ()

Protected Member Functions
void	check_mask_ (Layer< D > &layer, bool allow_oversized)
	Will check that the mask can be applied to the layer. More...

Protected Attributes
lockPTR< Ntree< D, index > >	ntree_

MaskDatum	mask_

Detailed Description

template<int D>
class nest::MaskedLayer< D >

Class for applying masks to layers.

Contains begin and end methods to iterate over nodes inside a mask.

Constructor & Destructor Documentation

template<int D>

nest::MaskedLayer< D >::MaskedLayer	(	Layer< D > &	layer,
		Selector	filter,
		const MaskDatum &	mask,
		bool	include_global,
		bool	allow_oversized
	)

inline

Regular constructor.

Parameters

layer	The layer to mask
filter	Optionally select subset of neurons
mask	The mask to apply to the layer
include_global	If true, include all nodes, otherwise only local to MPI process
allow_oversized	If true, allow larges masks than layers when using periodic b.c.

References nest::MaskedLayer< D >::check_mask_(), nest::Layer< D >::get_global_positions_ntree(), nest::Layer< D >::get_local_positions_ntree(), and nest::MaskedLayer< D >::ntree_.

template<int D>

nest::MaskedLayer< D >::MaskedLayer	(	Layer< D > &	layer,
		Selector	filter,
		const MaskDatum &	mask,
		bool	include_global,
		bool	allow_oversized,
		Layer< D > &	target
	)

inline

Constructor for applying "converse" mask to layer.

To be used for applying a mask for the target layer to the source layer. The mask will be mirrored about the origin, and settings for periodicity for the target layer will be applied to the source layer.

Parameters

layer	The layer to mask (source layer)
filter	Optionally select subset of neurons
mask	The mask to apply to the layer
include_global	If true, include all nodes, otherwise only local to MPI process
allow_oversized	If true, allow larges masks than layers when using periodic b.c.
target	The layer which the given mask is defined for (target layer)

References nest::MaskedLayer< D >::check_mask_(), nest::Layer< D >::get_extent(), nest::Layer< D >::get_global_positions_ntree(), nest::Layer< D >::get_lower_left(), nest::Layer< D >::get_periodic_mask(), nest::MaskedLayer< D >::mask_, and nest::MaskedLayer< D >::ntree_.

template<int D>

nest::MaskedLayer< D >::~MaskedLayer ( )

inline

Member Function Documentation

template<int D>

Ntree< D, index >::masked_iterator nest::MaskedLayer< D >::begin ( const Position< D > & anchor )

inline

Iterate over nodes inside mask.

Parameters

anchor Position to apply mask to

Returns: an iterator for the nodes inside the mask centered on the anchor position

References numerics::e.

Referenced by nest::ConnectionCreator::divergent_connect_(), nest::Layer< D >::get_global_nodes(), nest::Layer< D >::get_global_positions_vector(), and nest::ConnectionCreator::source_driven_connect_().

template<int D>

void nest::MaskedLayer< D >::check_mask_	(	Layer< D > &	layer,
		bool	allow_oversized
	)

protected

Will check that the mask can be applied to the layer.

The mask must have the same dimensionality as the layer, and a grid mask may only be applied to a grid layer. Unless the allow_oversized flag is set, the mask must also not be larger than the layer in case of periodic boundary conditions. Will throw an exception if the mask does not fit.

Parameters