`qml.fourier.coefficients` won't work with more than one expectation value

Issue description

When I try to calculate the Fourier coefficients with the new module qml.fourier if the circuit has more than one measurement an error appears.

Code:

import pennylane as qml
from pennylane import numpy as np
from pennylane.fourier import coefficients

dev = qml.device('default.qubit', wires=2)

@qml.qnode(dev)
def simple_circuit(x):
    qml.RX(x[0], wires=0)
    qml.RY(x[0], wires=1)
    qml.CNOT(wires=[1, 0])
    return qml.expval(qml.PauliZ(0)), qml.expval(qml.PauliZ(1))


coeffs = coefficients(simple_circuit, 1, 2)
print(np.round(coeffs, decimals=4))

Output:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
TypeError: only size-1 arrays can be converted to Python scalars

The above exception was the direct cause of the following exception:

ValueError                                Traceback (most recent call last)
<ipython-input-111-8c2ae78fdea3> in <module>
     13 
     14 
---> 15 coeffs = coefficients(simple_circuit, 1, 2)
     16 print(np.round(coeffs, decimals=4))

~/.local/lib/python3.8/site-packages/pennylane/fourier/coefficients.py in coefficients(f, n_inputs, degree, lowpass_filter, filter_threshold)
    154     """
    155     if not lowpass_filter:
--> 156         return _coefficients_no_filter(f, n_inputs, degree)
    157 
    158     if filter_threshold is None:

~/.local/lib/python3.8/site-packages/pennylane/fourier/coefficients.py in _coefficients_no_filter(f, n_inputs, degree)
    228 
    229         # fill discretized function array with value of f at inpts
--> 230         f_discrete[tuple(nvec)] = f(sample_points)
    231 
    232     coeffs = np.fft.fftn(f_discrete) / f_discrete.size

ValueError: setting an array element with a sequence.

[PCesteban]

Sure!, I will.

[mariaschuld]

@PCesteban let us know if you pick up anything else, your feedback is very valuable!