BUG: xarray groupby now leaving extra dimension

[michaelcdevin]

In wec.core.solve, the xarray groupby method is used to iterate over the different phase realizations. Previously, this has squeezed out the realization dimension in the DataArray and WecOptTool has indexed the wave information in the solve steps under that assumption.

However, squeezing the dimension was only occurring for monotonically increasing DataArrays, which led to confusion among xarray users and the functionality is no longer default and has been deprecated. As a result, as of xarray v2024.07.0, the dimensions are not squeezed, leaving the realization dimension in waves and causing downstream dimension sizing issues involving numpy.concatenate:

ValueError: all the input arrays must have same number of dimensions, but the array at index 0 has 2 dimension(s) and the array at index 1 has 3 dimension(s)

---------------
Traceback:
---------------
wecopttool\core.py:825: in solve
    optim_res = minimize(**problem)
C:\Users\mcdevin\AppData\Local\mambaforge\envs\wot_dev_new\Lib\site-packages\scipy\optimize\_minimize.py:743: in minimize
    res = _minimize_slsqp(fun, x0, args, jac, bounds,
C:\Users\mcdevin\AppData\Local\mambaforge\envs\wot_dev_new\Lib\site-packages\scipy\optimize\_slsqp_py.py:333: in _minimize_slsqp
    meq = sum(map(len, [atleast_1d(c['fun'](x, *c['args']))
C:\Users\mcdevin\AppData\Local\mambaforge\envs\wot_dev_new\Lib\site-packages\scipy\optimize\_slsqp_py.py:333: in <listcomp>
    meq = sum(map(len, [atleast_1d(c['fun'](x, *c['args']))
wecopttool\core.py:789: in scaled_resid_fun
    return self.residual(x_wec, x_opt, wave)
wecopttool\core.py:598: in residual
    ri = ri - f(self, x_wec, x_opt, waves)
wecopttool\core.py:2025: in force
    force_fd = complex_to_real(wave_excitation(force_coeff, waves), False)
wecopttool\core.py:1703: in complex_to_real
    out = np.concatenate([out, d])
C:\Users\mcdevin\AppData\Local\mambaforge\envs\wot_dev_new\Lib\site-packages\autograd\numpy\numpy_wrapper.py:38: in <lambda>
    concatenate = lambda arr_list, axis=0 : concatenate_args(axis, *arr_list)
C:\Users\mcdevin\AppData\Local\mambaforge\envs\wot_dev_new\Lib\site-packages\autograd\tracer.py:48: in f_wrapped
    return f_raw(*args, **kwargs)

[michaelcdevin]

Resolved in #365