Error is overestimated about fifipy HOT 2 CLOSED

Hi Dario,

I also looked at your cubik code and can only confirm Keith’s analysis.
It appears that it is not necessary to divide by the areafactor in line 372 of data.py. All the scaling should be taken care of in the expression above.
I’ve attached two screen shots of cubik with and without dividing by the areafactor. The graphical representation of the uncertainty in the screen shot without the division by areafactor looks more convincing than with the division. I chosen a radius of 2, to make the difference more apparent.

Please let us know, what’s the role of areafactor in the uncertainty calculation. Could it be that it is an obsolete remainder of some old code?

Best,
Randolf

from fifipy.

Change has been implemented

from fifipy.