Add option to pass a callback function to optimize()

[drpaprika]

This simple PR allows to pass a callback function to scipy's optimize().
This allows the user to add anything he likes to vizualize the optimization progress, for example (tested in a notebook):

import matplotlib.pyplot as plt
from IPython.display import clear_output

optimizer = optimization.OptimizerGeneric(problem)

history = []
def callback(_):
    # Plot the lens at each merit function call
    clear_output(wait=True)
    lens.draw()
    
    # Plot merit function evolution
    f_val = optimizer.problem.sum_squared()
    history.append(f_val)
    plt.figure(figsize=(6, 4))
    plt.plot(history, color='black')
    plt.yscale('log')
    plt.title("Merit function value")
    plt.xlabel("Iteration")
    plt.ylabel("Merit function value")
    plt.grid(alpha=0.25)
    plt.show()

res = optimizer.optimize(tol=1e-9, callback=callback)

Only LeastSquares doesn't seem to accept a callback function.

I also included a fix to update the lens at the end of the optimization with the best paramerters (result.x), as the last merit function call is not necessarily the lowest one.

Of course plotting during the optimization slows it significantly, but that's up to the user.
drpaprika

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

@@            Coverage Diff             @@
##           master      #79      +/-   ##
==========================================
- Coverage   96.57%   96.56%   -0.02%     
==========================================
  Files         105      105              
  Lines        6340     6348       +8     
==========================================
+ Hits         6123     6130       +7     
- Misses        217      218       +1     

Files with missing lines	Coverage Δ
optiland/optimization/optimization.py	98.17% <100.00%> (+0.03%)	⬆️

... and 1 file with indirect coverage changes

[HarrisonKramer]

Looks good. Thanks for the contribution!

And nice catch on updating to the best parameters after the optimization completes. I'll merge now.

Regards,
Kramer