[cctbxbb] MTZ object creation bug or feature?

[Gergely Katona]

Hi,

This is interesting and good to know. From F+/SIGF+,F-/SIGF one cannot
reconstruct DANO,SIGDANO if the errors are correlated and conversely
one loses information about correlations when going from DANO,SIGDANO
+ Fmean to F+/SIGF+,F-/SIGF . My preferred representation is therefore
DANO,SIGDANO + Fmean . It may be good not to make automatic
assumptions about these conversions.

Best wishes,

Gergely

On Tue, Mar 30, 2021 at 10:33 AM Huw Jenkins <h.t.jenkins at me.com> wrote:
>
> Hi Rob,
>
> > On 30 Mar 2021, at 01:27, Robert Oeffner <rdo20 at cam.ac.uk> wrote:
> >
> > Is this a bug, a well known feature or am I simply doing something wrong?
>
> This is certainly known (http://phenix-online.org/pipermail/phenixbb/2011-October/017772.html) - it's because when F,SIGF and DANO,SIGDANO are adjacent columns they are converted to F+/SIGF+,F-/SIGF- with Observation type: xray.reconstructed_amplitude
>
> Adding:
>
> for a in mtz1.mtz_object().as_miller_arrays():
>   if a.observation_type().__str__() == 'xray.reconstructed_amplitude':
>     for i,hkl in enumerate(a.indices()):
>       print('{:4d}{:4d}{:4d} {:0.2f} {:0.2f}'.format(hkl[0], hkl[1], hkl[2], a.data()[i], a.sigmas()[i]))
>
> to your example code shows this:
>
>    1  -2   3 12.30 4.02
>   -1   2  -3 4.10 4.02
>    0   0  -3 39.45 85.31
>    0   0   3 13.15 85.31
>    1   2   3 3.15 3.24
>   -1  -2  -3 1.05 3.24
>    0   1   2 7.35 68.42
>    0  -1  -2 2.45 68.42
>    1   0   2 48.60 1.23
>   -1   0  -2 16.20 1.23
>
>
> Best regards,
>
>
> Huw
> _______________________________________________
> cctbxbb mailing list
> cctbxbb at phenix-online.org
> http://phenix-online.org/mailman/listinfo/cctbxbb



-- 
Gergely Katona, Professor
Department of Chemistry and Molecular Biology, University of Gothenburg
Box 462, 40530 Göteborg, Sweden
Tel: +46-31-786-3959 / M: +46-70-912-3309 / Fax: +46-31-786-3910
Web: http://katonalab.eu, Email: gergely.katona at gu.se