The time has come to finally approach how the heck we're gonna handle uncertainties from the analysis.
Steps
Positional uncertainty from pads. This should be based purely on size of pad (probably triangle height?)
Positional uncertainty from TB. Idk exactly... We know what the precision of a single TB is I guess, so that could be a good starting point?
Propagate to the fit, and get fit errors. This is not natively supported by the L-BFGS-B method, but lmfit says it can with numdifftools. But maybe time to dust off leastsqs with bigger meshes... and maybe even secondary interpolation.
The time has come to finally approach how the heck we're gonna handle uncertainties from the analysis.
Steps