Check definition of prefactors/mixing orders

[jusack]

Currently calcPrefactors checks for amplitudes unequal to zero only in the second dimension of dfStrength()[1,:,1] which according to the MDF standard is the dimension of excitation coil channels at the first frequency applied on this channel. This assumes that the frequencies are ordered in a way that the primary component for each channel is the first frequency, which is the case for all measurements written by MPIMeasurements but as far as I can see is not required by the MDF standard. We should probably check the order of the frequencies to decide which are the main frequency components (diagonal or first column)

I have a patch in 4349d0b which can handle MDFs where the amplitudes are written on the diagonal, but this was just a quick fix for the is some details somewhat broken MDFs of the IMT 3D-MPS and I did not check any further if that is the correct way to define the function

[jusack]

Additionally it would be nice to be able to include non-drive field components into the calculations, e.g. shift or rotation frequencies. However this might need additional infrastructure on the MPIMeasurements side and/or custom MDF fields