Change modality parameter pixelsize_nm to take float

[tshakespeare]

Is your feature request related to a problem? Please describe.
User can change the pixel size and resolution of the modality they are using to match their experimental set-up, but the precision of the pixel size is limited to 1 nm, whereas in practise it should be ~0.01 nm. When comparing the size of structures from an imaging experiment to a vLab similation, the precision of the pixel size might cause avoidable (?) discrepencies. For example, I would like to compare nanocage size with those generated from simulations of nanocage labelling at varying efficiencies.

Describe the solution you'd like
Would prefer this variable to be a float, instead of an integer.

Describe alternatives you've considered
None