Discussion List Archives

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Comments on CIF core changes for 2.3

  • To: Multiple recipients of list <coredmg@iucr.org>
  • Subject: Re: Comments on CIF core changes for 2.3
  • From: "I. David Brown" <idbrown@mcmail.cis.mcmaster.ca>
  • Date: Mon, 14 Jul 2003 17:20:29 +0100 (BST)
Dear Curt,

	Thank you for your helpful comments on the draft of the proposed
changes to the cif_core dictionary.  I realize that you were not part of
this group for all the earlier discussions, but you have raised a number
of points that were not previously raised.

> data_atom_site_refinement_flags_adp
>   ...........
>     loop_ _enumeration
>           _enumeration_detail
>   .  'no constraints on atomic displacement parameters'
>   T   'special-position constraints on atomic displacement parameters'
>   U  'Uiso or Uij restraint (rigid bond)'
>   TU 'Both constraints applied'
>     _definition
> ;        A code which indicates the refinement restraints or constraints
>          applied to the atomic displacement parameters of this site.
> ;

The values of this flag were taken directly from the original core
dictionary.  I merely broke a single refinement_flag item into three
separate items, so there was no discussion about whether these flags were
appropriate.  My suggestion is that we approve this item, and discuss
additional flags during the discussions on version 2.4 in the autumn.

> ######################################################################
>
> data_atom_site_refinement_flags_occupancy
>      ..........
>     loop_ _enumeration
>           _enumeration_detail
        . 'no constraints on site occupancy parameters'
>       P 'site occupancy constraint'
>     _definition
> ;              A code which indicates that refinement restraints or
>                constraints were applied to the occupancy of this site.
> ;
>
> What is partial occupancy constraint?
> The most common restraint to occupancy is that the occupancies two partial
> occupancy sites sum to a constant (usually  1. but 0.5 etc for special
> position sites)  Also if the disorder is over three or more positions the
> sum of the occupancies can restrained to be a constant.

This flag is intended to warn a user that a constraint or restraint has
been applied to the occupancy of this atom.  Details of the constraint
should be given in the description of the refinement.  A partial occupancy
constraint is any constraint that prevents the free refinement of a
particular atomic occupancy.  Incidentally a fully occupied site, even if
it is a special position, has an occupancy of 1.0.  The restrictions
related to the symmetry of the special position appear in
_*_symmetry_multiplicity.

> ######################################################################
>
> data_diffrn_reflns_measured_fraction_resolution_full
    ...............
>      _related_item             '_diffrn_measured_fraction_theta_full'
>      _related_function          alternate
>      _definition
> ;         Fraction of unique (symmetry-independent) reflections measured
>           out to the resolution given in _diffrn_reflns_resolution_full.
>           This number should be very close to 1.0, since it represents the
>           fraction of reflections measured in the part of the diffraction
>           pattern that is essentially complete.
> ;
> # COMMENT: Replacement for _diffrn_measured_fraction_theta_full, moved to a
> # more appropriate category and defined in terms of resolution rather than
> # angle which depends on the radiation used.
> # 2003-06-26 _related_function 'replace' changed to 'alternate'
> # STATUS: Preliminary approval 2002-11-11
>
>
> There is the potential for confusion here.
> We currently have
> _diffrn_measured_fraction_theta_full
> _diffrn_measured_fraction_theta_max
> _diffrn_reflns_theta_full
> _diffrn_reflns_theta_max
> and we are adding
> _diffrn_reflns_measured_fraction_resolution_full
> _diffrn_reflns_measured_fraction_resolution_max
> _diffrn_reflns_resolution_full
> _diffrn_reflns_resolution_max
>
> In the experiment, I would think that
> _diffrn_reflns_measured_fraction_resolution_full (or _max)
> will be the same as the _diffrn_measured_fraction_theta_full (or _max).   I
> can see the value of talking in terms of resolution rather than theta,
> (although I am more comfortable with angles) but why have two data names
> that describe the same number.  Why not define terms
> _diffrn_reflns_measured_fraction_full and
> _diffrn_reflns_measured_fraction_max
> leaving out all reference to whether it is based on resolution or theta

This is a good idea.  I will remove the _*_measured_fraction_* names that
refer to resolution so that we can discuss introducing your suggested
names for revision 2.4.  In the mean time the 'theta_max' items can be
used.

> In addition the numbers currently being reported by a popular refinement
> program are based on what that refinement program is receiving as input, not
> what the data reduction program (integration program) had as input or
> output. Do we need or want entries similar to what currently exists in the
> dictionary for this situation.
> For example reflns_limit_ and diffrn_reflns_limit_ where the two values do
> not need to be the same as the first refers to the values used for
> refinement and the second to values in the data collection.
> Which are the more relevant numbers, the numbers related to refinement, or
> measurement, or both?

It is up to the writers of software and their users to make sure that they
write correct CIFs.  COMCIFS can point out errors in software when it
learns of them, and I suppose journals like Acta Cryst could make sure
that authors were aware of faults in their software.

The answer to your question about which resolution is more relevant, that
surely depends on why it is included.  In reporting a structure
determination I suppose the refinement resolution would be more important.

> ######################################################################
>
> data_diffrn_standards_decay_%_lt
   ...................
>     _related_item               '_diffrn_standards_decay_%'
>     _related_function           alternate
>     _definition
> ;          An upper limit on the percentage decrease in the mean
>            intensity of the set of standard reflections measured at the
>            start of the measurement of the diffraction pattern and at the
>            end. This value is used when the decay is too small to be
>            detected.
> ;
> # COMMENT: Many experiments show no detectable decay and there is no
> provision
> # for this at the moment other than to enter 0.0.
> # STATUS: Preliminary approval 2002-11-11
>
> I like the concept, but I wonder about its usage to create tables.  I have
> never written
> such programs, but seems complicated to me to have a table entry for crystal
> decay that looks at two different cif items to decide which one to use.
>
> I would guess that this item is being introduced because crystal decay is
> almost certainly never 0.0%.  Wouldn't it be simpler to change the
> definition of  '_diffrn_standards_decay_%' to reflect that decay corrections
> are approximate, not exact numbers.

Changing the definition would not indicate how approximate a value of 0.0
is.  On the other hand, Howard Flack suggest giving the decay a standard
uncertainty, which is certainly a valid way of indicating the limits on
how close to zero 0.0 is.

I will pull this definition from this draft and mark it for later
discussion.

> ######################################################################
>
>
> data__diffrn_reflns_measured_fraction_resolution_max
>
> There is a double __ here.

Thanks

> ######################################################################
>
>
> #
> # THE PREVIOUS ITEM REPLACES _symmetry_cell_setting.  The definition of the
> old
> # item should be modified to make it clear that the old name is discontinued
> # (the name is crystallographically incorrect and therefore misleading).
> The
> # following items should be added to the dictionary entry for
> # _symmetry_cell_setting:
> #
> #    _related_item        '_space_group_crystal_system'
> #    _related_function    replace
> #
> # Incidentally I have removed the _related_item and _related_function from
> the
> # definition of _space_group_crystal_system because they are incorrect.  It
> is
> # the _related_item that should 'replace' the item being defined, not the
> other
> # way around.
> #
> So what happens to _space_group_crystal_system??   Should there be a
> replaced or obsolete.

The DDL provides the opportunity to point out that a particular item has
been replaced by another item (or that it conveys the same information in
another form).  It does not, apparently, provide for a message to say that
the present item replaces an earlier and discontinued data item.  Thus
when a discontinued item that was originally defined in an earlier version
of the dictionary is also included (for completeness) in a later version
that includes its replacement, the presence of a _related_item field
points the user to the new name that should be used.  The DDL provides no
way to indicate that the present item replaces one that was previously
used to convey the same information.

> ######################################################################
>
>
>
> data_space_group_symop_operation_xyz
   ...........
>  loop_
>     _example
>     _example_detail
>       'x,1/2-y,1/2+z'  'c glide reflection through the plane (x,1/4,z)'
>     _definition
> ;      A parsable string giving one of the symmetry operations of the
>        space group in algebraic form.  If W is a matrix representation
>        of the rotational part of the symmetry operation defined by the
>        positions and signs of x, y and z, and w is a column of
>        translations defined by the fractions, an equivalent position
>                 X' is generated from a given position X by the equation:
>
>                           X' = WX + w
>
>        (Note: X is used to represent bold_italics_x in International
>         Tables for Crystallography Vol. A, Section 5)
>
>                 When a list of symmetry operations is given, it is assumed
>                 that the list contains all the operations of the space
>                 group (including the identity operation) as given by the
>                 representatives of the general position in International
>                 Tables for Crystallography Vol. A.
> ;
>     _type              char
>
>     _enumeration_default   'x,y,z'
>
> # STATUS: Preliminary approval 2003-03-11
>
> Doesn't c-glide define reflection and if so isn't reflection redundant?

Yes, but the explanation is not part of the definition, it is there just
for for illustration.  I will remove the offending word.

> I don't see how the example detail helps the definition.  I would like to
> see a fuller example but I do not know how to set it up to give something
> that looks like this.
>
> Thus for the symmetry operators for
> P212121 should be
> x,y,z
> -x,1/2+y,1/2-z
> 1/2-x,-y,1/2+z
> 1/2+x,1/2-y,-z
>
> and for C2/m (setting  ???? )
> x,y,z
> x,1/2-y,z
>
> or should it be
>
> and for C2/m (setting  ???? )
> x,y,z
> -x,-y,-z
> x,1/2-y,z
> -x,1/2+y,-z1/2+x,1/2+y,z
> 1/2-x,1/2-y,-z
> 1/2+x,-y,1/2+z
> 1/2-x,y,1/2-z
>
> I must admit that until I read (or misread) Howard's comments I would have
> made C2/m look like the simpler example.

Thanks for pointing out that the definition is ambiguous when it comes to
lattice translations.  In fact IT-A does include all the operations of
space group, but the lattice translations are implicit rather than
explicit.  I will clarify the wording to include lattice translations and
centers of symmetry. I suppose the thought was that eventually all
software would adopt the Hall symbol so that it would be unnecessary to
include this list.  However, the _geom fields refer to an ordered list of
symmetry operations and the Hall symbol does not specify a unique
ordering, so for the time being at least we need to continue with this
list.

> The suggested extended examples could help resolve ambiguities.  Perhaps the
> extended example could also include a non standard setting.

Examples do not need official approval and can be added as necessary.  I
will follow your suggestion and add one or two of the examples you
suggest.

				David


*****************************************************
Dr.I.David Brown,  Professor Emeritus
Brockhouse Institute for Materials Research,
McMaster University, Hamilton, Ontario, Canada
Tel: 1-(905)-525-9140 ext 24710
Fax: 1-(905)-521-2773
idbrown@mcmaster.ca
*****************************************************

[Send comment to list secretary]
[Reply to list (subscribers only)]