TileStepper.h

Classes

TileStepper -- traverse a tiled Lattice optimally with a tile cursor (full description)

class TileStepper: public LatticeNavigator

Interface

Public Members
TileStepper (const IPosition& latticeShape, const IPosition& tileShape)
TileStepper (const IPosition& latticeShape, const IPosition& tileShape, const IPosition& axisPath)
TileStepper (const TileStepper& other)
~TileStepper()
TileStepper& operator= (const TileStepper& other)
virtual Bool operator++(int)
virtual Bool operator--(int)
virtual void reset()
virtual Bool atStart() const
virtual Bool atEnd() const
virtual uInt nsteps() const
virtual IPosition position() const
virtual IPosition endPosition() const
virtual IPosition latticeShape() const
virtual IPosition subLatticeShape() const
virtual IPosition cursorShape() const
virtual IPosition cursorAxes() const
IPosition tileShape() const
virtual Bool hangOver() const
virtual void subSection (const IPosition& blc, const IPosition& trc)
virtual void subSection (const IPosition& blc, const IPosition& trc, const IPosition& inc)
virtual IPosition blc() const
virtual IPosition trc() const
virtual IPosition increment() const
virtual const IPosition& axisPath() const
virtual LatticeNavigator* clone() const
virtual Bool ok() const
Protected Members
virtual uInt calcCacheSize (const ROTiledStManAccessor&, uInt rowNumber) const
Private Members
TileStepper()

Description

Review Status

Reviewed By:
Peter Barnes
Date Reviewed:
1999/10/30
Programs:
Tests:

Prerequisite

Etymology

TileStepper is used to step optimally through a tiled Lattice.

Synopsis

When you wish to traverse a Lattice (say, a PagedArray or an Image) you will usually create a LatticeIterator. Once created, you may attach a LatticeNavigator to the iterator. A TileStepper is a concrete class derived from the abstract LatticeNavigator that allows you to step through the Lattice in a way that will minimize the amount of cache memory consumed and maximize the speed.

Some Lattices (in particular PagedArrays) are stored (on disk) in tiles. For an N-dimensional Lattice a tile is an N-dimensional subsection with fewer elements along each axis. For example a Lattice of shape [512,512,4,32] may have a tile shape of [32,16,4,16], and there will be 16*32*1*2 (=1024) tiles in the entire Lattice. To allow efficient access of the data in a Lattice some tiles are cached in memory. As each tile may consume a fair bit of memory (in this example 128kBytes, assuming each element consumes 4 bytes), it is desirable to minimise the number of tiles held in the cache. But it is also desirable to minimise the number of times a tiles must be read into or written from the cache as this may require a time consuming operation like disk I/O.

TileStepper steps through a lattice in a tile-by-tile way. This means that the cache contains 1 tile only and that a tile is accessed only once. It should be clear that traversing a lattice in this way cannot be used if an entire vector or plane is needed. It is, however, very well suited for purposes like initialising a lattice, where the order in which the lattice pixels are accessed is not important.

In constructing a TileStepper, you specify the Lattice shape, the tile shape and optionally the axis path. The axis path defines the order in which the tiles are fetched from the lattice. Default is the natural order (thus x-axis in the inner loop).
It is possible to use the function subSection to traverse only a subsection of the lattice.

The cursor position can be incremented or decremented to retrieve the next or previous tile in the Lattice. The position of the next tile in the Lattice will depend on the tile shape, and is described above.
Note that the cursor shape does not need to be constant when iterating through the lattice. If the lattice shape is not an integer multiple of the tile shape, the cursor will be smaller on the edges of the lattice.

Example

This example initializes a lattice with the given value.
    void init (Lattice<Complex>& cArray, Complex value)
    {
      const IPosition latticeShape = cArray.shape();
      const IPosition tileShape = cArray.niceCursorShape();
      TileStepper tsx(latticeShape, tileShape);
      LatticeIterator<Complex> lix(cArray, tsx);
      for (lix.reset();!lix.atEnd();lix++)
        lix.woCursor() = value;
      }
    }
    
Note that a TileStepper is the default navigator for an iterator. So the code above could be made simpler like shown below. Also note that this example is a bit artificial, because the Lattice::set() function should be used to initialize a lattice.
    void init (Lattice<Complex>& cArray, Complex value)
    {
      LatticeIterator<Complex> lix(cArray);
      for (lix.reset();!lix.atEnd();lix++)
        lix.woCursor() = value;
      }
    }
    

Motivation

This class makes it possible to traverse a lattice in the optimal way.

Member Description

TileStepper (const IPosition& latticeShape, const IPosition& tileShape)
TileStepper (const IPosition& latticeShape, const IPosition& tileShape, const IPosition& axisPath)

Construct a TileStepper by specifying the Lattice shape, a tile shape, and an optional axis path (default is natural order). Is is nearly always advisable to make the tileShape identical to the Lattice tileShape. This can be obtained by lat.niceCursorShape() where lat is a Lattice object.

TileStepper (const TileStepper& other)

Copy constructor (copy semantics).

~TileStepper()

TileStepper& operator= (const TileStepper& other)

Assignment (copy semantics).

virtual Bool operator++(int)

Increment operator (postfix or prefix version) - move the cursor forward one step. Returns True if the cursor was moved.

virtual Bool operator--(int)

Decrement operator (postfix or prefix version) - move the cursor backwards one step. Returns True if the cursor was moved.

virtual void reset()

Function to move the cursor to the beginning of the Lattice. Also resets the number of steps (nsteps function) to zero.

virtual Bool atStart() const

Function which returns "True" if the cursor is at the beginning of the Lattice, otherwise, returns "False"

virtual Bool atEnd() const

Function which returns "True" if an attempt has been made to increment the cursor beyond the end of the Lattice.

virtual uInt nsteps() const

Function to return the number of steps (increments & decrements) taken since construction (or since last reset). This is a running count of all cursor movement (operator++ or operator--), even though N-increments followed by N-decrements will always leave the cursor in the original position.

virtual IPosition position() const

Function which returns the current position of the beginning of the cursor. The position function is relative to the origin in the main Lattice.

virtual IPosition endPosition() const

Function which returns the current position of the end of the cursor. The endPosition function is relative the origin in the main Lattice.

virtual IPosition latticeShape() const
virtual IPosition subLatticeShape() const

Functions which return the shape of the Lattice being iterated through. latticeShape always returns the shape of the main Lattice while subLatticeShape returns the shape of any sub-Lattice defined using the subSection function.

virtual IPosition cursorShape() const

Function which returns the shape of the cursor. This always includes all axes (i.e. it includes degenerates axes)

virtual IPosition cursorAxes() const

Function which returns the axes of the cursor.

IPosition tileShape() const

Function which returns the shape of the "tile" the cursor will iterate through before moving onto the next tile.

virtual Bool hangOver() const

Function which returns "True" if the increment/decrement operators have moved the cursor position such that part of the cursor beginning or end is hanging over the edge of the Lattice. This always returns False.

virtual void subSection (const IPosition& blc, const IPosition& trc)
virtual void subSection (const IPosition& blc, const IPosition& trc, const IPosition& inc)

Functions to specify a "section" of the Lattice to step over. A section is defined in terms of the Bottom Left Corner (blc), Top Right Corner (trc), and step size (inc), on ALL of its axes, including degenerate axes. The step size defaults to one if not specified.

virtual IPosition blc() const
virtual IPosition trc() const
virtual IPosition increment() const

Return the bottom left hand corner (blc), top right corner (trc) or step size (increment) used by the current sub-Lattice. If no sub-Lattice has been defined (with the subSection function) these functions return blc=0, trc=latticeShape-1, increment=1, ie. the entire Lattice.

virtual const IPosition& axisPath() const

Return the axis path.

virtual LatticeNavigator* clone() const

Function which returns a pointer to dynamic memory of an exact copy of this instance. The pointer returned by this function must be deleted externally.

virtual Bool ok() const

Function which checks the internal data of this class for correct dimensionality and consistant values. Returns True if everything is fine otherwise returns False

virtual uInt calcCacheSize (const ROTiledStManAccessor&, uInt rowNumber) const

Calculate the cache size (in tiles) for this type of access to a lattice in the given row of the tiled hypercube.

TileStepper()

Prevent the default constructor from being used.