flopscope.numpy.can_cast
fnp.can_cast(*args, **kwargs)[flopscope source]
Returns True if cast between data types can occur according to the casting rule.
Adapted from NumPy docs np.can_cast
Cost
0
Flopscope Context
Returns True if cast is safe.
Parameters
- from_:dtype, dtype specifier, NumPy scalar, or array
Data type, NumPy scalar, or array to cast from.
- to:dtype or dtype specifier
Data type to cast to.
- casting:{'no', 'equiv', 'safe', 'same_kind', 'unsafe'}, optional
Controls what kind of data casting may occur.
'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.
Returns
- out:bool
True if cast can occur according to the casting rule.
See also
Notes
Changed in version 2.0.
Examples
Basic examples
>>> import flopscope.numpy as fnp
>>> flops.can_cast(flops.int32, flops.int64)
True
>>> flops.can_cast(flops.float64, complex)
True
>>> flops.can_cast(complex, float)
False>>> flops.can_cast('i8', 'f8')
True
>>> flops.can_cast('i8', 'f4')
False
>>> flops.can_cast('i4', 'S4')
False