non-integer data

ckegan

Say I had a diffraction pattern that had intensity values that had non-integer intensities ranging from 0 -> 5 (e.g. take a normal pattern and divide by 1000 - this is not what I actually have but demonstrates the point), what effect would this have on the calculated R values or on the quantities to be refined?

Could I then recover 'normal' values by multiplying by 1000 and rounding to integers? Do I risk getting rounding errors?

Thanks

Chris

johnsoevans

If you rescale a powder pattern you have to be careful to rescale the errors as well.