Error passing significant digits and two numpy array containing strings

[SimonRubenDrauz]

Describe the bug
I have two objects I want to compare with each other. In order to do so I use your package deepdiff. So imagine you have object A and object B. In order to check, if both objects are equal, I did the following:

DeepDiff(self._dict_, other._dict_, ignore_nan_inequality=True, significant_digits=6, math_epsilon=1e-6,
ignore_private_variables=False).

Doing this in case of a list of strings in _dict_, I get the following error:

File "C:...\lib\site-packages\numpy\testing_private\utils.py", line 1039, in compare z = abs(x - y)
TypeError: unsupported operand type(s) for -: 'str' and 'str'.

A quickfix that worked for me in self._diff_numy_array:

If you agree on that quickfix, I can do a PR, if you want to.

To Reproduce

a = np.array(['a', 'b'], dtype=object)
b = np.array(['a', 'b'], dtype=object)
DeepDiff(a, b, significant_digits=6)

Expected behavior
I would expect that passing a dictionary where significant_digits apply to only few entries, in case of the other entries it is just skipped.

OS, DeepDiff version and Python version (please complete the following information):

• OS: Windows
• Version: 10 Enterprise
• Python Version: 3.8.5
• DeepDiff Version: 6.1.0

Additional context
The numpy version I use is 1.22.4

[seperman]

Hi @SimonRubenDrauz
Thank you for reporting this issue and providing the solution too. Sorry it took more than a year to take a look at this ticket.
I copy pasted your solution and pushed it to the dev branch. I will keep you posted once it is released.