Hi Tim,


I was discussing that with a student today, because we observe the same thing 
when we make test sets for k-fold cross validation. It is an implementation 
choice. You can random sort and then bin, or you can assign a random number to 
each reflection and translate that to a bin. The latter is fine for large bins, 
but not for small ones like the ones you are making.



Cheers,

Robbie







Sent from my Windows 10 phone



Van: Tim Gruene<mailto:[email protected]>
Verzonden: maandag 30 januari 2017 21:03
Aan: [email protected]<mailto:[email protected]>
Onderwerp: [ccp4bb] variation in freerflag



Dear all,

when I ran freerflag with the keyword FREERFRAC 0.002, the number of  reflection
per flag (between 0 and 500) is very high, it varies between 17 and 1.

Why is there such a high variation, instead of flagging always the same number
of reflections per flag (6-7 in my case)? Can the distribution be made evenly?

Best regards,
Tim
--
--
Paul Scherrer Institut
Tim Gruene
- persoenlich -
OFLC/102
CH-5232 Villigen PSI
phone: +41 (0)56 310 5297

GPG Key ID = A46BEE1A

Reply via email to