flopscope.numpy.copyto

fnp.copyto(dst, src, casting='same_kind', where=True)[flopscope source]

Copies values from one array to another, broadcasting as necessary.

Adapted from NumPy docs np.copyto

Areacore

Typecustom

Cost

per-operation

Flopscope Context

Copy values from src to dst array. Cost: num copied.

Raises a TypeError if the casting rule is violated, and if where is provided, it selects which elements to copy.

Parameters

dst:ndarray: The array into which values are copied.
src:array_like: The array from which values are copied.
casting:{'no', 'equiv', 'safe', 'same_kind', 'unsafe'}, optional: Controls what kind of data casting may occur when copying.
'no' means the data types should not be cast at all.
'equiv' means only byte-order changes are allowed.
'safe' means only casts which can preserve values are allowed.
'same_kind' means only safe casts or casts within a kind, like float64 to float32, are allowed.
'unsafe' means any data conversions may be done.
where:array_like of bool, optional: A boolean array which is broadcasted to match the dimensions of dst, and selects elements to copy from src to dst wherever it contains the value True.

>>> import flopscope.numpy as fnp
>>> A = flops.array([4, 5, 6])
>>> B = [1, 2, 3]
>>> flops.copyto(A, B)
>>> A
array([1, 2, 3])

>>> A = flops.array([[1, 2, 3], [4, 5, 6]])
>>> B = [[4, 5, 6], [7, 8, 9]]
>>> flops.copyto(A, B)
>>> A
array([[4, 5, 6],
       [7, 8, 9]])