javadba / nptyping

Type hints for Numpy

nptyping

Type hints for Numpy!

Installation

pip install nptyping

Usage

Use the nptyping type hints like the regular type hints from typing:

from nptyping import Array


def func1(arr: Array[int]):  # A numpy.ndarray that contains numbers
    ...

You can also define the shape of an array:

Array[str, 3, 2]    # 3 rows and 2 columns
Array[str, 3]       # 3 rows and an undefined number of columns
Array[str, 3, ...]  # 3 rows and an undefined number of columns
Array[str, ..., 2]  # an undefined number of rows and 2 columns

Heterogeneous arrays are supported as well:

Array[int, float, str]       # int, float and str on columns 1, 2 and 3 resp.
Array[int, float, str, ...]  # int, float and str on columns 1, 2 and 3 resp.
Array[int, float, str, 3]    # 3 rows and int, float and str on columns 1, 2 and 3 resp.

nptyping also supports instance checks:

import numpy
from nptyping import Array


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

isinstance(arr, Array[int, 3, 2])    # True
isinstance(arr, Array[str, 3, 2])    # False
isinstance(arr, Array[int, 3, ...])  # True
isinstance(arr, Array[int, 3, 6])    # False

Also for heterogeneous arrays:

import numpy
from nptyping import Array


arr = np.array([(1, 2.0, '3'),
                (4, 5.0, '6')],
               dtype=[('a', int), ('b', float), ('c', str)])

isinstance(arr, Array[int, float, str])     # True
isinstance(arr, Array[float, float, str])   # False
isinstance(arr, Array[int, float, str, 2])  # True