7. Structured arrays wrapping

Taking derivatives on structured numpy arrays is not supported by jax, so structured arrays are internally wrapped with StructuredArray.

7.1. StructuredArray

class lsqfitgp.StructuredArray(array)

JAX-friendly imitation of a numpy structured array.

It behaves like a read-only numpy structured array, and you can create a copy with a modified field with a jax-like syntax.

Parameters:

arraynumpy array, StructuredArray

A structured array. An array qualifies as structured if array.dtype.names is not None.

Notes

The StructuredArray is a readonly view on the input array. When you change the content of a field of the StructuredArray, however, the reference to the original array for that field is lost.

Examples

>>> a = numpy.empty(3, dtype=[('f', float), ('g', float)])
>>> a = StructuredArray(a)
>>> a = a.at['f'].set(numpy.arange(3))
is equivalent to
>>> a['f'] = numpy.arange(3)

classmethod from_dataframe(df)

Make a Structured array from a DataFrame. Data is not copied.

7.2. Functions

lsqfitgp.asarray(x, dtype=None)

Version of numpy.asarray that works with StructuredArray and JAX arrays. If x is not an array already, returns a JAX array if possible.

lsqfitgp.broadcast(*arrays)

Version of numpy.broadcast that works with StructuredArray.

lsqfitgp.broadcast_arrays(*arrays, **kw)

Implementation of numpy.broadcast_arrays for lgp.StructuredArray.

Broadcast any number of arrays against each other.

Parameters:

`*args`array_likes

The arrays to broadcast.

subokbool, optional

If True, then sub-classes will be passed-through, otherwise the returned arrays will be forced to be a base-class array (default).

Returns:

broadcastedlist of arrays

These arrays are views on the original arrays. They are typically not contiguous. Furthermore, more than one element of a broadcasted array may refer to a single memory location. If you need to write to the arrays, make copies first. While you can set the writable flag True, writing to a single output value may end up changing more than one location in the output array.

Deprecated since version 1.17: The output is currently marked so that if written to, a deprecation warning will be emitted. A future version will set the writable flag False so writing to it will raise an error.

See also

broadcast

broadcast_to

broadcast_shapes

Examples

>>> x = np.array([[1,2,3]])
>>> y = np.array([[4],[5]])
>>> np.broadcast_arrays(x, y)
[array([[1, 2, 3],
       [1, 2, 3]]), array([[4, 4, 4],
       [5, 5, 5]])]

Here is a useful idiom for getting contiguous copies instead of non-contiguous views.

>>> [np.array(a) for a in np.broadcast_arrays(x, y)]
[array([[1, 2, 3],
       [1, 2, 3]]), array([[4, 4, 4],
       [5, 5, 5]])]

lsqfitgp.broadcast_to(x, shape, **kw)

Implementation of numpy.broadcast_to for lgp.StructuredArray. Broadcast an array to a new shape.

Parameters:

arrayarray_like

The array to broadcast.

shapetuple or int

The shape of the desired array. A single integer i is interpreted as (i,).

subokbool, optional

If True, then sub-classes will be passed-through, otherwise the returned array will be forced to be a base-class array (default).

Returns:

broadcastarray

A readonly view on the original array with the given shape. It is typically not contiguous. Furthermore, more than one element of a broadcasted array may refer to a single memory location.

Raises:

ValueError

If the array is not compatible with the new shape according to NumPy’s broadcasting rules.

See also

broadcast

broadcast_arrays

broadcast_shapes

Notes

New in version 1.10.0.

Examples

>>> x = np.array([1, 2, 3])
>>> np.broadcast_to(x, (3, 3))
array([[1, 2, 3],
       [1, 2, 3],
       [1, 2, 3]])

lsqfitgp.unstructured_to_structured(arr, dtype=None, names=None, align=False, copy=False, casting='unsafe')

Like numpy.lib.recfunctions.unstructured_to_structured, but outputs a StructuredArray.