flattenFrom

Return a copy of an input ndarray where all dimensions of the input ndarray are flattened starting from a specified dimension.

Usage

var flattenFrom = require( '@stdlib/ndarray/flatten-from' );

flattenFrom( x, dim[, options] )

Returns a copy of an input ndarray where all dimensions of the input ndarray are flattened starting from a specified dimension.

var array = require( '@stdlib/ndarray/array' );

var x = array( [ [ [ 1.0, 2.0 ] ], [ [ 3.0, 4.0 ] ], [ [ 5.0, 6.0 ] ] ] );
// returns <ndarray>

var y = flattenFrom( x, 1 );
// returns <ndarray>[ [ 1.0, 2.0 ], [ 3.0, 4.0 ], [ 5.0, 6.0 ] ]

The function accepts the following arguments:

• x: input ndarray. Must have one or more dimensions.
• dim: dimension to start flattening from. If provided an integer less than zero, the dimension index is resolved relative to the last dimension, with the last dimension corresponding to the value -1.
• options: function options (optional).

The function accepts the following options:

• order: order in which input ndarray elements should be flattened. Must be one of the following:

  • 'row-major': flatten elements in lexicographic order. For example, given a two-dimensional input ndarray (i.e., a matrix), flattening in lexicographic order means flattening the input ndarray row-by-row.
  • 'column-major': flatten elements in colexicographic order. For example, given a two-dimensional input ndarray (i.e., a matrix), flattening in colexicographic order means flattening the input ndarray column-by-column.
  • 'any': flatten according to the physical layout of the input ndarray data in memory, regardless of the stated order of the input ndarray.
  • 'same': flatten according to the stated order of the input ndarray.

  Default: 'row-major'.

• dtype: output ndarray data type. By default, the function returns an ndarray having the same data type as a provided input ndarray.

By default, the input ndarray is flattened in lexicographic order. To flatten elements in a different order, specify the order option.

var array = require( '@stdlib/ndarray/array' );

var x = array( [ [ [ 1.0, 2.0 ] ], [ [ 3.0, 4.0 ] ], [ [ 5.0, 6.0 ] ] ] );
// returns <ndarray>

var y = flattenFrom( x, 0, {
    'order': 'column-major'
});
// returns <ndarray>[ 1.0, 3.0, 5.0, 2.0, 4.0, 6.0 ]

By default, the output ndarray data type is inferred from the input ndarray. To return an ndarray with a different data type, specify the dtype option.

var array = require( '@stdlib/ndarray/array' );
var dtype = require( '@stdlib/ndarray/dtype' );

var x = array( [ [ [ 1.0, 2.0 ] ], [ [ 3.0, 4.0 ] ], [ [ 5.0, 6.0 ] ] ] );
// returns <ndarray>

var y = flattenFrom( x, 0, {
    'dtype': 'float32'
});
// returns <ndarray>[ 1.0, 2.0, 3.0, 4.0, 5.0, 6.0 ]

var dt = String( dtype( y ) );
// returns 'float32'

Notes

• The function always returns a copy of input ndarray data, even when an input ndarray already has the desired number of dimensions.

Examples

var discreteUniform = require( '@stdlib/random/array/discrete-uniform' );
var array = require( '@stdlib/ndarray/array' );
var ndarray2array = require( '@stdlib/ndarray/to-array' );
var flattenFrom = require( '@stdlib/ndarray/flatten-from' );

var xbuf = discreteUniform( 12, -100, 100, {
    'dtype': 'generic'
});

var x = array( xbuf, {
    'shape': [ 2, 2, 3 ],
    'dtype': 'generic'
});
console.log( ndarray2array( x ) );

var y = flattenFrom( x, 1 );
console.log( ndarray2array( y ) );