MultiArray

Convert an array of Ruby objects to a multi-dimensional Sequence_ object.  The element type of the returned sequence is choosen to fit all elements.

Parameters

Returns

Object of class Sequence_

def MultiArray.import( type, data, * shape )

Create a multi-dimensional array from raw data.

Parameters

def MultiArray.load_raw( file )

Construct from file created with Sequence_.save_raw.

See also

• Sequence_.save_raw

def MultiArray.new( element_type, * shape )

Construct new multi-dimensional array with elements of same type.  The elements will not be initialized.  Use fill! if you need initialized elements.

Parameters

def MultiArray.object( * shape )

Shortcut for MultiArray.new( OBJECT, *shape )

See also

• MultiArray.new

def MultiArray.bool( * shape )

Shortcut for MultiArray.new( BOOL, *shape )

See also

• MultiArray.new

def MultiArray.ubyte( * shape )

Shortcut for MultiArray.new( UBYTE, *shape )

See also

• MultiArray.new

def MultiArray.byte( * shape )

Shortcut for MultiArray.new( BYTE, *shape )

See also

• MultiArray.new

def MultiArray.usint( * shape )

Shortcut for MultiArray.new( USINT, *shape )

See also

• MultiArray.new

def MultiArray.sint( * shape )

Shortcut for MultiArray.new( SINT, *shape )

See also

• MultiArray.new

def MultiArray.uint( * shape )

Shortcut for MultiArray.new( UINT, *shape )

See also

• MultiArray.new

def MultiArray.int( * shape )

Shortcut for MultiArray.new( INT, *shape )

See also

• MultiArray.new

def MultiArray.ulong( * shape )

Shortcut for MultiArray.new( ULONG, *shape )

See also

• MultiArray.new

def MultiArray.long( * shape )

Shortcut for MultiArray.new( LONG, *shape )

See also

• MultiArray.new

def MultiArray.dfloat( * shape )

Shortcut for MultiArray.new( DFLOAT, *shape )

See also

• MultiArray.new

def MultiArray.sfloat( * shape )

Shortcut for MultiArray.new( SFLOAT, *shape )

See also

• MultiArray.new

def MultiArray.dcomplex( * shape )

Shortcut for MultiArray.new( DCOMPLEX, *shape )

See also

• MultiArray.new

def MultiArray.scomplex( * shape )

Shortcut for MultiArray.new( SCOMPLEX, *shape )

See also

• MultiArray.new

def MultiArray.ubytergb( * shape )

Shortcut for MultiArray.new( UBYTERGB, *shape )

See also

• MultiArray.new

def MultiArray.bytergb( * shape )

Shortcut for MultiArray.new( BYTERGB, *shape )

See also

• MultiArray.new

def MultiArray.usintrgb( * shape )

Shortcut for MultiArray.new( USINTRGB, *shape )

See also

• MultiArray.new

def MultiArray.sintrgb( * shape )

Shortcut for MultiArray.new( SINTRGB, *shape )

See also

• MultiArray.new

def MultiArray.uintrgb( * shape )

Shortcut for MultiArray.new( UINTRGB, *shape )

See also

• MultiArray.new

def MultiArray.intrgb( * shape )

Shortcut for MultiArray.new( INTRGB, *shape )

See also

• MultiArray.new

def MultiArray.ulongrgb( * shape )

Shortcut for MultiArray.new( ULONGRGB, *shape )

See also

• MultiArray.new

def MultiArray.longrgb( * shape )

Shortcut for MultiArray.new( LONGRGB, *shape )

See also

• MultiArray.new

def MultiArray.dfloatrgb( * shape )

Shortcut for MultiArray.new( DFLOATRGB, *shape )

See also

• MultiArray.new

def MultiArray.sfloatrgb( * shape )

Shortcut for MultiArray.new( SFLOATRGB, *shape )

See also

• MultiArray.new

def MultiArray.laplacian_of_gaussian( sigma, size )

Compute Laplacian of gaussian filter.  No error estimate is given.  The LoG-filter will display zero-crossings at edges.  The edges can be detected using a combination of dilation and erosion.  Alternatively one can match a set of small 3x3 patterns.  Neither technique is provided by this library at the moment.

See also

• Laplacian of Gaussian

def MultiArray.load_ubyte( file )

Load a 8-bit greyscale-image.  A colour image will be converted to greyscale.

Parameters

def MultiArray.load_usint( file )

Load a 16-bit greyscale-image.  A colour image will be converted to greyscale.

Parameters

def MultiArray.load_ubytergb( file )

Load a colour-image.  A greyscale image will be represented with RGB values.

Parameters

def MultiArray.load_sfloat( file )

Load a single-precision floating-point greyscale-image.  A colour image will be converted to greyscale.

Parameters

def MultiArray.load_dfloat( file )

Load a double-precision floating-point greyscale-image.  A colour image will be converted to greyscale.

Parameters

def MultiArray.load_sfloatrgb( file )

Load a single-precision floating-point colour-image.  A greyscale image will be represented with RGB values.

Parameters

def MultiArray.load_dfloatrgb( file )

Load a double-precisino floating-point colour-image.  A greyscale image will be represented with RGB values.

Parameters

def MultiArray.load( file )

Load image using RMagick.  It is recommended to use one of the load_* methods instead.

See also

• <loadf>

def MultiArray.loadf( file )

Load image using OpenEXR.  It is recommended to use one of the load_* methods instead.

See also

• <load>

def MultiArray.tensor( n )

Computes a tensor product for an n-dimensional result.  This method expects a code block as argument.  The code block is interpreted using special objects as arguments.  Therefore only certain operations are possible.  The implementation is inspired by Walter Landry’s FTensor library.  The following example shows a matrix multiplication as an example

v = MultiArray.sint( 2, 3 ).indgen!
w = MultiArray.sint( 3, 2 ).indgen!
r = MultiArray.tensor( 2 ) { |i,j,k| v[k,j] * w[i,k] }

Parameters

See also

• http://www.oonumerics.org/FTensor/