Skip to main content

Type hints for Numpy.

Project description

PyPI version Downloads PyPI version codecov Scrutinizer Code Quality

Type hints for Numpy!

(❒) Installation

pip install nptyping

(❒) Usage

(❒) NDArray

nptyping.NDArray lets you define the shape and type of your numpy.ndarray.

You can:

  • specify the number of dimensions;
  • specify the size per dimension;
  • specify the type of the array;
  • instance check your array with your nptying type.

(❒) Examples

An Array with any dimensions of any size and any type:

>>> from nptyping import NDArray
>>> from typing import Any


>>> NDArray
NDArray[(typing.Any, ...), typing.Any]

>>> NDArray[(Any, ...)]
NDArray[(typing.Any, ...), typing.Any]

>>> NDArray[(Any, ...), Any]
NDArray[(typing.Any, ...), typing.Any]

An array with 1 dimension of any size and any type:

>>> NDArray[Any]
NDArray[(typing.Any,), typing.Any]

>>> NDArray[(Any,)]
NDArray[(typing.Any,), typing.Any]

>>> NDArray[Any, Any]
NDArray[(typing.Any,), typing.Any]

>>> NDArray[(Any,), Any]
NDArray[(typing.Any,), typing.Any]

An array with 1 dimension of size 3 and any type:

>>> NDArray[3]
NDArray[(3,), typing.Any]

>>> NDArray[(3,)]
NDArray[(3,), typing.Any]

>>> NDArray[(3,), Any]
NDArray[(3,), typing.Any]

An array with 3 dimensions of size 3, 3 and any and any type:

>>> NDArray[3, 3, Any]
NDArray[(3, 3, typing.Any), typing.Any]

>>> NDArray[(3, 3, Any)]
NDArray[(3, 3, typing.Any), typing.Any]

>>> NDArray[(3, 3, Any), Any]
NDArray[(3, 3, typing.Any), typing.Any]

An array with any dimensions of any size and type int:

>>> import numpy as np

>>> NDArray[np.int32]
NDArray[(typing.Any, ...), Int[32]]

>>> NDArray[(Any, ...), np.int32]
NDArray[(typing.Any, ...), Int[32]]

Note that provided types are translated to nptyping types. Pure Python types (.e.g int or float are supported as well). You can also provide nptyping types yourself: NDArray[(Any, ...), Int[64]].

An array with 1 dimension of size 3 and type int:

>>> NDArray[3, np.int32]
NDArray[(3,), Int[32]]

>>> NDArray[(3,), np.int32]
NDArray[(3,), Int[32]]

An array with any dimensions of size 3 and type int:

>>> NDArray[(3, ...), np.int32]
NDArray[(3, ...), Int[32]]

An array with 3 dimensions of sizes 3, 3, 5 and type int:

>>> NDArray[(3, 3, 5), np.int32]
NDArray[(3, 3, 5), Int[32]]

A structured array:

>>> import numpy as np

>>> NDArray[(Any,...), np.dtype([('x',np.int32), ('y',np.int32)])]
NDArray[(typing.Any, ...), StructuredType[Int[32], Int[32]]]

(❒) Checking your instances

You can use NDArray with isinstance to dynamically check your arrays.

>>> import numpy as np

>>> arr = np.array([[1, 2, 3],
...                 [4, 5, 6]])

>>> isinstance(arr, NDArray[(2, 3), int])
True
>>> isinstance(arr, NDArray[(2, 3), float])
False
>>> isinstance(arr, NDArray[(2, 3, 1), int])
False

(❒) Finding the right annotation

You can use NDArray to find the type of a numpy array for you using NDArray.type_of:

>>> NDArray.type_of(np.array([[1, 2], [3, 4.0]]))
NDArray[(2, 2), Float[64]]

See also nptyping.get_type (documented below).

(❒) Int

An nptyping equivalent of numpy signed integers.

>>> from nptyping import Int

>>> Int[32]
Int[32]

You can also use one of these:

>>> from nptyping import Int8, Int16, Int32, Int64

(❒) UInt

An nptyping equivalent of numpy unsigned integers.

>>> from nptyping import UInt

>>> UInt[64]
UInt[64]

You can also use one of these:

>>> from nptyping import UInt8, UInt16, UInt32, UInt64

(❒) Float

An nptyping equivalent of numpy floats.

>>> from nptyping import Float

>>> Float[64]
Float[64]

You can also use one of these:

>>> from nptyping import Float16, Float32, Float64

(❒) Unicode

An nptyping equivalent of numpy unicodes.

>>> from nptyping import Unicode

>>> Unicode[100]
Unicode[100]

(❒) Bool

An nptyping equivalent of numpy bool.

>>> from nptyping import Bool

>>> Bool
Bool

(❒) Complex128

An nptyping equivalent of numpy complex128.

>>> from nptyping import Complex128

>>> Complex128
Complex128

(❒) Datetime64

An nptyping equivalent of numpy datetime64.

>>> from nptyping import Datetime64

>>> Datetime64
Datetime64

(❒) Timedelta64

An nptyping equivalent of numpy timedelta64.

>>> from nptyping import Timedelta64

>>> Timedelta64
Timedelta64

(❒) Object

An nptyping equivalent of numpy objects.

>>> from nptyping import Object

>>> Object
Object

(❒) StructuredType

An nptyping equivalent of numpy structured dtypes.

>>> from nptyping import StructuredType, Int

>>> StructuredType[Int[32], Int[32]]
StructuredType[Int[32], Int[32]]

(❒) SubArrayType

An nptyping equivalent of numpy subarray dtypes.

>>> from nptyping import SubArrayType, Int

>>> SubArrayType[Int[16], (4,2)]
SubArrayType[Int[16], (4, 2)]

(❒) get_type

With get_type you can get nptyping equivalent types for your arguments:

>>> from nptyping import get_type

>>> get_type(np.int32)
Int[32]
>>> get_type('some string')
Unicode[11]
>>> get_type(np.dtype([('x', np.int32), ('y', np.int32)]))
StructuredType[Int[32], Int[32]]

(❒) py_type

With py_type you can get the Python builtin type that corresponds to a Numpy dtype:

>>> from nptyping import py_type

>>> py_type(np.int32)
<class 'int'>

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for nptyping, version 1.4.0
Filename, size File type Python version Upload date Hashes
Filename, size nptyping-1.4.0-py3-none-any.whl (30.0 kB) File type Wheel Python version py3 Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page