Old XDS file formats
Jump to navigation
Jump to search
For some purposes it may be useful to use the following INPUT_FILE formats mentioned in the XSCALE documentation :
ANOMAL.HKL
(ASCII, formatted sequential)
Description of file format ANOMAL as produced by XDS (ANOMAL.HKL)
or XSCALE (XSCALE.HKL) for the compact representation of anomalous
intensity data. This file-format will be automatically selected by
XDS and XSCALE if a positive value for the input parameter DELFRM
is specified (XDS.INP) or >Frame separation (XSCALE.INP).
All reflections that are related by crystallographi symmetry are
averaged and saved in a single record FORMAT(3I5,8E12.4)
Each record consists of the following items
h,k,l,IwP,SDwP,IwM,SDwM,IP,SDP,IM,SDM
h,k,l - unique reflection indices. The file is sorted with
respect to these indices.
h=10000 indicates the last record in file.
IwP,SDwP - Mean weighted intensity and its standard deviation
of reflections strictly symmetry related to h, k, l
IwM,SDwM - Mean weighted intensity and its standard deviation
of reflections strictly symmetry related to -h,-k,-l
IP,SDP - Mean unweighted intensity and its standard deviation
of reflections strictly symmetry related to h, k, l
IM,SDM - Mean unweighted intensity and its standard deviation
of reflections strictly symmetry related to -h,-k,-l
The unweighted intensities are computed from a
restricted set of reflections. Only those reflections
are included which form a BIJVOET pair recorded on
images separated by less than DELFRM (see XDS.INP)
images in the data set.
DEFINITIONS and COMMENTS
A *negative* standard deviation indicates that the intensity has not
been measured.
A *zero* standard deviation indicates that -h,-k,-l is strictly
symmetry related to h, k, l.
Mean *weighted* intensity = SUMi{Ii/SDi**2}/SUMi{1/SDi**2}
Mean *unweighted* intensity = SUMi{Ii}/SUMi{1}
For computation of the mean *unweighted* intensity,
reflection #i is included in the summation only
if there exists a different reflection #j such
that i,j forms a Bijvoet pair and
|image_number(#i) - image_number(#j)| < DELFRM
NORMAL.HKL
(ASCII, formatted sequential)
Description of file format NORMAL (or OLDHKL) as produced by XDS
(NORMAL.HKL) and XSCALE (XSCALE.HKL) for representing unique
intensities in the absence of anomalous scattering effects.
Symmetry related reflections including Friedel-pairs are averaged
and recorded by FORTRAN FORMAT(3I5,4E12.4).
Each record consists of the following items
h,k,l,I,SDI
h,k,l - unique reflection indices.
h=10000 indicates the last record in file.
I,SDI - Mean weighted intensity and its standard deviation
of all reflections symmetry related to h, k, l or
-h,-k,-l
In case SDI is missing on an input file of this type,
XSCALE will assume SDI=0.1*I.
UNIQUE.HKL
(formatted sequential)
This file format has been replaced by either ANOMAL or NORMAL, but
files of this type may still be used as a reference data set.
Symmetry related reflections are averaged and written with
FORMAT(3I5,4E12.4). Each record consists of
HA,KA,LA,I,Sigma(I),DI,Sigma(DI)
HA,KA,LA - unique reflection indices. The file is sorted with
respect to these indices. HA=10000 indicates the
last record in file.
I,Sigma(I) - Mean intensity and its standard deviation.
DI,Sigma(DI) - Anomalous intensity difference and its standard
deviation. Missing data are indicated by Sigma(DI)<0.
DEFINITIONS and COMMENTS
1)Sigma(DI)>0 No missing data
-----------
Anomalous scattering effects are possible for HA, KA, LA
and expected (as indicated by the input parameter DELFRM>0).
I = ('Iw+' + 'Iw-')/2 ; Sigma(I) = standard deviation of I
DI = 'I+' - 'I-' ; Sigma(DI)= standard deviation of DI
'Iw-' weighted mean of all reflection intensities which are
strictly symmetry related to -HA,-KA,-LA;
'Iw+' weighted mean of all reflection intensities which are
strictly symmetry related to HA, KA, LA;
'I+' are the unweighted means of reflections included in the
'I-' estimation of Bijvoet pairs as controlled by the input
parameter DELFRM>0.
NOTE: 'I+' and 'I-' together with their standard deviations
can be recovered in good approximation as :
'I+' = I + DI/2 ; 'I-' = I - DI/2;
Sigma('I+')=Sigma('I-')=Sigma(DI)/sqrt(2)
which is sufficient information for obtaining Bayesian estimates
of structure factor amplitudes and their anomalous scattering
differences. This is carried out in the scaling routine XSCALE.
2)Sigma(DI)=0 No missing data
-----------
Anomalous scattering effects cannot occur for HA, KA, LA or
are negligable (as indicated by the input parameter DELFRM<=0).
I = weighted mean of all reflection intensities
which have the same unique indices HA,KA,LA;
Sigma(I) = standard deviation of I
DI = 0 because there is no anomalous scattering effect
Sigma(DI)= 0 standard deviation of DI is theoretically 0.0
3)Sigma(DI)<0 Some data are missing in this record.
----------- The value of DI indicates which data are missing.
a) DI<0: missing data 'Iw+' and DI.
I = 'Iw-' ; Sigma(I) = standard deviation of 'Iw-'
b) DI>0: missing data 'Iw-' and DI.
I = 'Iw+' ; Sigma(I) = standard deviation of 'Iw+'
c) DI=0: missing only data DI. Although both 'Iw+' and 'Iw-'
exist, this case may happen because of the value of
the input parameter DELFRM which controls acceptance
of Bijvoet pairs.
I =('Iw+' + 'Iw-')/2; Sigma(I) = standard deviation of I