Test failures with numpy 2.2

[manivmtc]

Bug

Tests fail due to numpy version specific imports

_________ ERROR collecting tests/utils/structures/test_key_line_2d.py __________  
ImportError while importing test module '/home/DGP/dgp/tests/utils/structures/test_key_line_2d.py'.  
Hint: make sure your test modules/packages have valid Python names.  
Traceback:  
../../../../miniconda3/envs/envDGP/lib/python3.10/importlib/__init__.py:126: in import_module  
    return _bootstrap._gcd_import(name[level:], package, level)  
tests/utils/structures/test_key_line_2d.py:3: in <module>  
    from numpy.lib.ufunclike import fix  
E   ModuleNotFoundError: No module named 'numpy.lib.ufunclike'

Same issue with dgp/tests/utils/structures/test_key_line_3d.py and dgp/tests/utils/test_camera.py

To Reproduce

Steps to reproduce the behavior:

1. Install numpy version 2.2

Minimum reproducible code example:
Just create a virtual environment using pyenv or conda. Activate it and then try this on the command-line
make test
Also, you can do
python -c "import numpy; print(numpy.__version__); from numpy.lib import ufunclike"

Expected behavior

If the problem is corrected, the tests should pass. The imports should no longer fail

Additional context

RHEL machine, so cannot install numpy 1.20.3 using pip for some weird gcc compatibility reasons

[YoungseokOh]

install numpy 1.23

_dgp 1.0 requires numpy>=1.20.3, but you have numpy 1.20.0 which is incompatible.
contourpy 1.3.0 requires numpy>=1.23, but you have numpy 1.20.0 which is incompatible.
matplotlib 3.8.4 requires numpy>=1.21, but you have numpy 1.20.0 which is incompatible.
pandas 2.2.3 requires numpy>=1.22.4; python_version < "3.11", but you have numpy 1.20.0 which is incompatible.
xarray 2024.7.0 requires numpy>=1.23, but you have numpy 1.20.0 which is incompatible._

numpy 1.20.3 is incompatible.

so, it works well for the test.

pip install numpy==1.23