Update use of numpy's random number generators.

[jcapriot]

Allow the use of numpy's default random number generator in discretize. Still a bit of work to do to make it complete though.

[codecov]

Codecov Report

Attention: Patch coverage is 97.27891% with 8 lines in your changes missing coverage. Please review.

Project coverage is 86.28%. Comparing base (98caca7) to head (186130e).
Report is 21 commits behind head on main.

Files with missing lines	Patch %	Lines
discretize/utils/mesh_utils.py	0.00%	3 Missing ⚠️
examples/plot_dc_resistivity.py	0.00%	3 Missing ⚠️
examples/plot_cahn_hilliard.py	0.00%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #360      +/-   ##
==========================================
+ Coverage   86.25%   86.28%   +0.03%     
==========================================
  Files          89       89              
  Lines       18695    18717      +22     
  Branches     2959     2961       +2     
==========================================
+ Hits        16125    16150      +25     
+ Misses       1884     1881       -3     
  Partials      686      686              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[jcapriot]

Believe this is ready to go now. Only thing to discuss is the name of the keyword. I put it as rng instead of seed to make it clearer that you can pass a random number generator and not just a seed.

The other thing is that I made it so you can't easily set the random number generator for the success/fail messages. I think it's more interesting to keep getting different random messages. :)

[santisoler]

Looks good to me! I noticed that this PR also replaces the pymatsolver solvers for the scipy ones. I think it's ok if you want to leave those changes in this PR as long as we add it to the list of changes in the description and in the commit message.