search

PyEllips / pyElli

An open source ellipsometry analysis tool for reproducible and comprehensible building of optical models.
https://pyelli.readthedocs.io
GNU General Public License v3.0
19 stars 7 forks source link

Support for measurement uncertainties #94

Open MarJMue opened 2 years ago

MarJMue commented 2 years ago

While thinking about #93 i got the idea, that it might be usefull to add support for measurement uncertainties. I'm not sure if lmfit / scipy.optimize can use this information ootb, but pint (#91) can and we could combine it with #53.

domna commented 2 years ago

I think lmfit does support calculating confidence intervals for a fit instead of the error estimates but not really using an uncertainty of the measurement data. I don't know if this can be somehow injected in the minimize function lmfit is using, but other than that we could also do three fits on the mean and edge case data and calculate the uncertainty from this (but well.... not my favourite idea :D). scipy however has a parameter to insert a std dev, at least I used it some time ago.

I think it's a nice feature, because afaik is the measurement uncertainty included in files generated from Woollam instruments