It's based on an assumption of Gaussian errors, but it is certainly arguable 
about whether this is the best weighting scheme

Phil


On 18 Mar 2010, at 17:40, Jacob Keller wrote:

> Okay, I think I understand:
> 
> <I> is the weighted average of the equivalent scaled intensities, with 
> 1/sigma^2 as weight for each reflection.
> 
> Is there a derivation/rationale for using 1/sigma^2 as the weight? I see that 
> this would give more credence to better I/sigma measurements, but I do not 
> yet understand why this is exactly the correct weight. Why not simply 
> 1/sigma, or 1/sigma^87?
> 
> I apologize for the basic nature of this inquiry, but I would like to 
> understand all of this fully. And...you can always ignore!
> 
> Jacob
> 
> *******************************************
> Jacob Pearson Keller
> Northwestern University
> Medical Scientist Training Program
> Dallos Laboratory
> F. Searle 1-240
> 2240 Campus Drive
> Evanston IL 60208
> lab: 847.491.2438
> cel: 773.608.9185
> email: [email protected]
> *******************************************
> 
> ----- Original Message ----- From: "Jim Pflugrath" <[email protected]>
> To: <[email protected]>
> Sent: Thursday, March 18, 2010 10:12 AM
> Subject: Re: [ccp4bb] Equation for combining measurements
> 
> 
>> The weighted mean is used.  Wikipedia
>> http://en.wikipedia.org/wiki/Weighted_mean has the equation(s).
>> 
>> 
>> -----Original Message-----
>> From: CCP4 bulletin board [mailto:[email protected]] On Behalf Of
>> Eleanor Dodson
>> Sent: Thursday, March 18, 2010 4:59 AM
>> To: [email protected]
>> Subject: Re: [ccp4bb] Equation for combining measurements
>> 
>> Isnt it in the scala documentation?
>> Eleanor
>> 
>> Jacob Keller wrote:
>>> Dear Crystallographers,
>>> 
>>> A basic question: what is the equation currently used to combine/merge
>>> multiple measurements of I and sigI? (Or a reference would be fine as
>>> well (or maybe better.))
>>> 
>>> Thanks,
>>> 
>>> Jacob Keller
>>> 
>>> *******************************************
>>> Jacob Pearson Keller
>>> Northwestern University
>>> Medical Scientist Training Program
>>> Dallos Laboratory
>>> F. Searle 1-240
>>> 2240 Campus Drive
>>> Evanston IL 60208
>>> lab: 847.491.2438
>>> cel: 773.608.9185
>>> email: [email protected]
>>> ******************************************* 

Reply via email to