How to apply numpy's integer array indexing both to ndarray-s and python lists?

Question

Numpy has "integer array indexig", which allows to index ndarray with other ndarray or list:

>> A = np.arange(10,20)
>> A[[1,2,3]]
Out[14]: array([11, 12, 13])

Suppose I don't know what is A, which can be either ndarray or Python's list.

Is there any explicit indexing function in numpy, which would allow the same indexing and accept both types?

For example:

>> A = np.arange(10,20)
>> np.get_elements(A, [1,2,3])
Out[14]: array([11, 12, 13])
>> A = range(10,20)
>> np.get_elements(A, [1,2,3])
Out[15]: [11, 12, 13]

Do you expect it to return a list if input is a list? A numpy function will first ensure A is an array (eg asarday(A)). — hpaulj
– hpaulj, Commented Nov 20, 2017 at 15:37

hpaulj · Accepted Answer · 2017-11-20 17:35:39Z

1

def get_elements(A, idx):
    try:
        return A[idx]
    except TypeError:
        import operator
        return list(operator.itemgetter(*idx)(A))

In [35]: get_elements(np.arange(10), [1,3,4])
Out[35]: array([1, 3, 4])
In [36]: get_elements(np.arange(10).tolist(), [1,3,4])
Out[36]: [1, 3, 4]

itemgetter is just a convenience class. The list comprehension would be just as good

In [39]: [A[i] for i in [1,3,4]]
Out[39]: [1, 3, 4]

A numpy function like take returns an array

In [40]: np.take(A,[1,3,4])
Out[40]: array([1, 3, 4])

which could be converted back into a list if desired.

answered Nov 20, 2017 at 17:35

hpaulj

233k14 gold badges260 silver badges392 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How to apply numpy's integer array indexing both to ndarray-s and python lists?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related