[buster-discuss] aP_scale for multiple crystals

[Francis Reyes]

Hi all,

I'm reading https://www.globalphasing.com/autoproc/wiki/index.cgi?ExampleUsage#combine with regard to performing scaling using multiple crystals. 

I have pointless output (from mosflm runs) of each crystal. I use combine_files to ensure consistent indexing and to output a single mtz with (in this case) three batches corresponding to three crystals. 

The steps after this on the wiki:
[1] Use a constant scale factor and specifying each batch separately. 
[2] Try smooth scale factors as a function of rotation "ROTATION SPACING" etc. 


In both cases, by specifying each batch separately, what is happening?
Are the scale factors determined by ensuring [1] corresponding reflections in each dataset are similar AND [2] that their symmetry related reflections within each dataset are similar? 

At the end of this run, I obtain three separate files (each corresponding to different batches). However, I would like a single file consisting of the most complete dataset ? Would I specify -b 1000-4000? 

Thanks,

F