Statistics on Calibrations

[AnnaWiniwarter]

For validating calibration curves, it could be a good idea to add some info on the quality of the linear fit, i.e. a possibility of making a plot of residuals vs. concentration or similar. And if we're getting very fancy also adding some confidentiality intervals. dreaming big

(note, this issue was transferred here from the ixdat issues page number 106)

[ScottSoren]

A start of this is taken care of by SensitivityUnion, which is the result when adding two SensitivityFactors with the same mol and mass:

spectro_inlets_quantification/src/spectro_inlets_quantification/sensitivity.py

Line 150 in b595ba4

class SensitivityUnion(SensitivityFactor):

It acts like a normal SensitivityFactor but has an attribute F_error which is the relative standard deviation of the sensitivity factor values.