Allow complex numbers in pints.vector()

[HOLL95]

The pints.vector(array) command copies the input array with dtype=float, which casts the imaginary components from a complex input to real.

[MichaelClerx]

We'll have to think a bit about what allowing complex numbers (and other number types?) will affect the different error measures and likelihoods.

One easy way would be to do a quick check in their constructors and just raise an exception if the number type is not a supported one. But even that's not straightforward, because we'd have to check e.g. convertibility-to-float rather than isinstance(float) so that we can allow other implementations of floats etc.