0

I would like to write a python function as below:

import numpy as np
a = [[-0.17985, 0.178971],[-0.15312,0.226988]] 
(lambda x: x if x > 0 else np.exp(x)-1)(a)

Below is python error message:

TypeError                                 Traceback (most recent call last)
<ipython-input-8-78cecdd2fe9f> in <module>
----> 1 (lambda x: x if x > 0 else np.exp(x)-1)(a)

<ipython-input-8-78cecdd2fe9f> in <lambda>(x)
----> 1 (lambda x: x if x > 0 else np.exp(x)-1)(a)

TypeError: '>' not supported between instances of 'list' and 'int'

How do I fix this problem?

For example:

a = [[-0.17985, 0.178971],[-0.15312,0.226988]]

b = f(a)

expected output

b = [[-0.1646, 0.17897],[-0.14197, 0.22699]]
Amir Shabani
  • 3,857
  • 6
  • 30
  • 67
user2286499
  • 3
  • 2
  • '>' not supported between instances of 'list' and 'int' - nobody knows why you are comparing that (supposedly matrix) list with 0. But it is what raises the error. – dmitry Apr 06 '19 at 03:02
  • Thanks to give me response. Can you guide me how to write a python function to do element-wise operation if element x > 0 return x; otherwise, take exp(x)-1 – user2286499 Apr 06 '19 at 04:08

1 Answers1

1

You are having a list of lists, so an additional iteration is required:

import numpy as np

a = [[-0.17985, 0.178971],[-0.15312,0.226988]]

f = lambda x: x if x > 0 else np.exp(x)-1

res = []
for x in a:
    lst = []
    for y in x:
        lst.append(f(y))
    res.append(lst)

print(res)
# [[-0.16460448865975663, 0.17897099999999999], [-0.14197324757693675, 0.226988]]

Since end result is a list, this problem can better be solved using a list-comprehension:

[[x if x > 0 else np.exp(x)-1 for x in y] for y in a]

Or with the defined lambda:

[[f(x) for x in y] for y in a]
Austin
  • 25,759
  • 4
  • 25
  • 48