[cctbxbb] unmerged data as mtz object

[Robert Oeffner]

Thank you David. That looks useful to me.

Rob

On 04-08-2020 18:49, David Waterman wrote:
> You might like to look at the MTZ export functionality in DIALS
> https://github.com/dials/dials/blob/master/util/export_mtz.py
>
> -- David
>
>
> On Tue, 4 Aug 2020 at 18:45, Robert Oeffner <rdo20 at cam.ac.uk 
> <mailto:rdo20 at cam.ac.uk>> wrote:
>
>     Thanks Aaron, I was aware of the  merge_equivalents function. My
>     interest is just for doing some testing on my code which involves
>     reading and writing unmerged data. I guess I could always use the
>     iotbx.mtz functions and write individual mtz columns including
>     separate HKL indices to a file.
>
>     Rob
>
>
>     On 04-08-2020 18:03, Aaron Brewster wrote:
>>     Hi Rob, I don't know how to create an unmerged mtz file, but if
>>     you want to merge the data you can use the cctbx miller array
>>     function merge_equivalents:
>>
>>     mtzobj =
>>     millarr.merge_equivalents().as_mtz_dataset(column_root_label="I")
>>
>>     I believe it does simple weighted averaging of the data using the
>>     sigmas.
>>
>>     -Aaron
>>
>>     On Tue, Aug 4, 2020 at 6:16 AM Robert Oeffner <rdo20 at cam.ac.uk
>>     <mailto:rdo20 at cam.ac.uk>> wrote:
>>
>>         Hi,
>>
>>         Does anybody know how to convert a miller array of unmerged
>>         data to an mtz object? For merged data I do something like this:
>>
>>         mtzobj = millarr.as_mtz_dataset(column_root_label="I")
>>
>>         But if millarr contains multiple observations with the same
>>         hkl index. Then the above code fails with the exception
>>
>>         cctbx Error: Duplicate entries in miller_indices array.
>>
>>         Any suggestions?
>>
>>         Many thanks,
>>
>>         Rob
>>



-- 
This email has been checked for viruses by AVG.
https://www.avg.com
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://phenix-online.org/pipermail/cctbxbb/attachments/20200804/c137abe9/attachment.htm>