larray.Array.expand¶

Array.expand(self, target_axes=None, out=None, readonly=False)[source]¶

Expands array to target_axes.

Target axes will be added to array if not present. In most cases this function is not needed because LArray can do operations with arrays having different (compatible) axes.

Parameters:

target_axes : string, list of Axis or AxisCollection, optional: Self can contain axes not present in target_axes. The result axes will be: [self.axes not in target_axes] + target_axes
out : Array, optional: Output array, must have more axes than array. Defaults to a new array. arr.expand(out=out) is equivalent to out[:] = arr
readonly : bool, optional: Whether returning a readonly view is acceptable or not (this is much faster)

Returns:

Array: Original array if possible (and out is None).

Examples

>>> a = Axis('a=a1,a2')
>>> b = Axis('b=b1,b2')
>>> arr = ndtest([a, b])
>>> arr
a\b  b1  b2
 a1   0   1
 a2   2   3

Adding one or several axes will append the new axes at the end

>>> c = Axis('c=c1,c2')
>>> arr.expand(c)
 a  b\c  c1  c2
a1   b1   0   0
a1   b2   1   1
a2   b1   2   2
a2   b2   3   3

If you want to new axes to be inserted in a particular order, you have to give that order

>>> arr.expand([a, c, b])
 a  c\b  b1  b2
a1   c1   0   1
a1   c2   0   1
a2   c1   2   3
a2   c2   2   3

But it is enough to list only the added axes and the axes after them:

>>> arr.expand([c, b])
 a  c\b  b1  b2
a1   c1   0   1
a1   c2   0   1
a2   c1   2   3
a2   c2   2   3