PyTaco Data Type Object

class pytaco.dtype

A tensor contains elements describe by this dtype object.

PyTaco currently does not provide a way to construct your own data types but provides several common data types for users.

See also

as_np_dtype

Convert to NumPy dtype

Notes

PyTaco exports the following data types:

pytaco.bool - A True or False value.

pytaco.int8 - An 8 bit signed integer.

pytaco.int16- A 16 bit signed integer.

pytaco.int32 - A 32 bit signed integer.

pytaco.int64 - A 64 bit signed integer.

pytaco.uint8 - An 8 bit unsigned integer.

pytaco.uint16 - A 16 bit unsigned integer.

pytaco.uint32 - A 32 bit unsigned integer.

pytaco.uint64 - A 64 bit unsigned integer.

pytaco.float32 or pytaco.float - A 32 bit floating point number.

pytaco.float64 or pytaco.double - A 64 bit floating point number.

PyTaco also overrides the equality operator of the data type class so users can compare types using == and != to check if they are the same.

Examples

>>> import pytaco as pt
>>> pt.int32.is_uint()
False
>>> pt.uint32.is_int()
False
>>> pt.float32 == pt.float64
False
>>> pt.bool != pt.int64
True
>>> pt.int8
pytaco.int8_t

Methods

is_bool(self)	Returns True if the data type is a boolean type and False otherwise.
is_uint(self)	Returns True if the data type is an unsigned integer and False otherwise.
is_int(self)	Returns True if the data type is a signed integer and False otherwise.
is_float(self)	Returns True if the data type is a float or double and False otherwise.