Some nice juxtapositions today. After a couple of days trying to get literal values working properly, I finally sorted out my rusty statistics and trigonometry enough to derive magnitudes and orientations from images. That meant that I could build a first function that is parameterised by those. More to come tomorrow.
But meanwhile, a spurious picture generated at some point during this morning's experimentation with those vectors (they may not have been right at this point - mistakenly using a figure in degrees where radians are expected takes a lot longer to diagnose than the opposite).