modified logging to use python lists rather than preallocated arrays

[kamdh]

This is more pythonic and there was a bug with previous preallocation. The info is still returned as a numpy array. I've attached a minimal example that reproduces the error with the saved data.

Error:

NpzFile 'temp_data.npz' with keys: theta, y
Linesearch failed with error 1. Damping max BB scaling to 10000.0
Linesearch failed with error 1. Damping max BB scaling to 1000.0
Linesearch failed with error 1. Damping max BB scaling to 100.0
Linesearch failed with error 1. Damping max BB scaling to 10.0
Linesearch failed with error 1. Damping max BB scaling to 1.0
Linesearch failed with error 1. Damping max BB scaling to 0.1
Traceback (most recent call last):
  File "/home/kameron/src/Visual-Compressive-Sensing/err_ex_spgl1.py", line 10, in <module>
    s, r, g, info = spg_bp(A, b, iter_lim=10240)
  File "/home/kameron/src/Visual-Compressive-Sensing/.venv/lib/python3.10/site-packages/spgl1/spgl1.py", line 1495, in spg_bp
    x, r, g, info = spgl1(A, b, tau, sigma, x0, **kwargs)
  File "/home/kameron/src/Visual-Compressive-Sensing/.venv/lib/python3.10/site-packages/spgl1/spgl1.py", line 1189, in spgl1
    xnorm1[niters] = primal_norm(x, weights)
IndexError: index 10240 is out of bounds for axis 0 with size 10240

min_err_example.zip