Function that is agnostic to object implementing len() or shape

user_4764124 0
Viewed 52

I have a function func(x) which expects a parameter x. x should be some kind of collection or array. func(x) needs to get the length of x:

 def func(x):        
    for i in range(len(x)):
        print('Something', i)

The problem is that some collection-like objects have len() implemented but not .shape (e.g. a list) others have .shape implemented but not len() (e.g. a sparse matrix). What is the most pythonic way to make func(x) work, regardless of this issue? Or are there other things to consider when looking for a solution for this problem?

What I can think of:

  • Expecting another parameter: func(x, length_of_x)

  • Calling a different function altogether

  • Catching missing implementation with try-except:

     import numpy as np
 from scipy.sparse import csr_matrix
 def func(x):
     try:
         length_of_x = len(x)
     except:
         length_of_x = x.shape[0]

     for i in range(length_of_x):
         print('Something', i)

 func([1, 2, 3])

 sparse_matrix = csr_matrix(([1, 2, 3, 4, 5, 6], ([0, 0, 1, 2, 2, 2], [0, 2, 2, 0, 1, 2])))
 func(sparse_matrix)

  • if-else Block

     def func(x):
     if isinstance(x, csr_matrix):
         length_of_x = x.shape[0]
     else:
         length_of_x = len(x)   
     for i in range(length_of_x):
         print('Something', i)
function oop python
1 Answers

The most Pythonic way to approach would be to build the internal logic using iterators rather than indexes.

def func(x):        
    for value in x:
        print('Something', value)
user_52375600
Related