mlx.core.issubdtype

issubdtype(: mlx.core.Dtype, : mlx.core.Dtype) → bool

issubdtype(: mlx.core.Dtype, : mlx.core.DtypeCategory) → bool

issubdtype(: mlx.core.DtypeCategory, : mlx.core.Dtype) → bool

issubdtype(: mlx.core.DtypeCategory, : mlx.core.DtypeCategory) → bool

Check if a Dtype or DtypeCategory is a subtype of another.

>>> ints = mx.array([1, 2, 3], dtype=mx.int32)
>>> mx.issubdtype(ints.dtype, mx.integer)
True
>>> mx.issubdtype(ints.dtype, mx.floating)
False

>>> floats = mx.array([1, 2, 3], dtype=mx.float32)
>>> mx.issubdtype(floats.dtype, mx.integer)
False
>>> mx.issubdtype(floats.dtype, mx.floating)
True

Similar types of different sizes are not subdtypes of each other:

>>> mx.issubdtype(mx.float64, mx.float32)
False
>>> mx.issubdtype(mx.float32, mx.float64)
False

but both are subtypes of floating:

>>> mx.issubdtype(mx.float64, mx.floating)
True
>>> mx.issubdtype(mx.float32, mx.floating)
True

For convenience, dtype-like objects are allowed too:

>>> mx.issubdtype(mx.float32, mx.inexact)
True
>>> mx.issubdtype(mx.signedinteger, mx.floating)
False