[Date Prev][Date Next][Date Index]
(54) Observer; CIF Core 2.0; CIF Core 2.1
- To: [email protected]
- Subject: (54) Observer; CIF Core 2.0; CIF Core 2.1
- From: bm
- Date: Wed, 27 Nov 1996 14:17:38 GMT
Dear Colleagues
I welcome Professor Andre Authier as the Observer from the Commission on
Crystallographic Nomenclature to our discussion list.
A54.1 CIF Core version 2.0
--------------------------
I have frozen the working draft of 27 November 1996 as the release version
of the CIF Core Dictionary version 2.0. It is available by anonymous ftp
at the URL
ftp://ftp.iucr.ac.uk/pub/cif_core_2.0.dic
A PostScript version of the formatted dictionary will be available from the
CIF home page, and will be mailed to you in the near future.
The last changes made in fine-tuning the dictionary before release are
summarised below.
Thank you all for your valuable input into this version of the dictionary.
If at times the discussions have seemed unduly protracted, I apologise; but
it has seemed important to have the core dictionary which underpins all the
extension dictionaries well supported by this Committee as a whole.
----------------------------------------------------------------------------
1996-11-14 BMcM:
Imposed consistency on the nomenclature of diffraction device
data names: _diffrn_detector_device_type -> *_detector_type,
_diffrn_measurement_device_details -> *_measurement_details,
*_measurement_device_type -> *_measurement_type; introduction
of _diffrn_source_device and parallel definitions. Existing
*_measurement_details example moved to *_special_details (PMDF) 1996-11-21 BMcM:
Reintroduced _diffrn_measurement_device_details, further tidying
of data names thus: _diffrn_measurement_type -> *_device_type;
_diffrn_detector_device and _diffrn_source_device both drop
"_device" (PMDF)
Added _journal_data_validation_number and
_publ_requested_category to enable handling of CIF-access
submissions by Acta Cryst. C
1996-11-23 A few typos fixed. BMcM
1996-11-24 BMcM:
Added 'gaussian', 'multi-scan' and 'numerical' to enumeration
list for _exptl_absorpt_correction_type (SRH)
Added 'mixed' to enumeration list for
_refine_ls_hydrogen_treatment (SRH)
1996-11-25 A few typos fixed (BMcM)
1996-11-27 BMcM:
Removed looped _related_item from _publ_contact_author and
reintroduced _diffrn_radiation_detector, *_dtime and
_diffrn_radiation_source (see 1996-11-08) with
"_related_function replace" as a preparation for
using this mechanism further in version 2.1.
1996-11-27 Release version 2.0. IUCr
----------------------------------------------------------------------------
===========================================
D54.1 Proposal for handling future versions
-------------------------------------------
It is apparent that the timescale for future updates is likely to be less
than 5 years (see below). I am therefore suggesting a modification to our
approach for the administration of the dictionaries. As new proposals for
additional data names are approved, they shall be entered into a draft
version of the dictionary mounted as a public "beta" distribution. Such beta
versions will be open to community review (along the lines of the successful
mmCIF process). I intend also to open a discussion list (by email server,
WWW bulletin board or some combination) for CIF software developers, which
will permit advance notice and discussion of changes with software authors.
The CIF Home Page will contain links to both the most recent supported
dictionary (2.0, as of now), and the most recent beta release. Beta versions
will be considered as relatively stable, but with the possibility of changes
arising from demonstrable error. It is likely that a number of beta
modifications will accumulate before another "full release" is announced.
For example, I have some proposals from Syd arising from the evolving needs
of Acta C. They include (a) replacement of some data names by new ones
better expressing current nomenclature and practice, (b) additional data
names to resolve the remaining problems with calculated wR and S, and (c)
a description of area detector experiments. If these were to attract
increasing amounts of debate, one might envisage versions 2.1beta1, 2.1beta2
and 2.1beta3 incorporating progressively (a), (b) and (c), before the final
release of version 2.1.
To give you an idea of how this will work, I am going to introduce these
points in the next sections. My hope is that they will attract no controversy,
and the requested changes can be put in place by 12 December, so that Syd's
Notes for Authors which goes to press then can refer to the new data names
as listed in version 2.1beta1 of the core.
(Note that not every minor change need lead immediately to a new beta release.)
D54.2 Replacement data names
----------------------------
Under pressure from the Commission on Journals to eradicate the term
'observed' to describe measurements that fall outside some specified
threshold, Syd proposes a number of changes in data name. The way this must
work in practice, to maintain compatibility with older files, is to add to
the existing definition a set of attributes
_related_item '_whatever'
_related_function replace
that point to the new definition (to see how this works, look at the
entry for _diffrn_radiation_source in the new version 2.0 file). The names
listed in Notes for Authors (new and old names) are
_reflns_number_gt Number of reflections > sigma threshold (new)
_reflns_number_observed Number of "observed" reflections
_reflns_threshold_sigma_expression
Sigma expression for F, F2 or I threshold (new)
_reflns_observed_criterion Sigma expression for "observed" F, F2 or I
_refine_ls_R_factor_gt R factor of F for reflections > threshold (new)
_refine_ls_R_factor_obs R factor of F for "observed" reflections
For completeness, the following data names referring to 'obs' should also be
replaced:
_refln_observed_status
_refine_ls_goodness_of_fit_obs
_refine_ls_restrained_S_obs
_refine_ls_wR_factor_obs
_reflns_shell_meanI_over_sigI_obs
_reflns_shell_number_measured_obs
_reflns_shell_number_unique_obs
_reflns_shell_percent_possible_obs
_reflns_shell_Rmerge_F_obs
_reflns_shell_Rmerge_I_obs
There is also pressure from the Nomenclature Commission to drop the use of
the term e.s.d. and replace it by standard uncertainty (this has already
been implemented in the descriptions within the Dictionary). Hence Syd
proposes:
_refine_ls_shift/su_max
Maximum shift/su ratio after final refinement cycle (new)
_refine_ls_shift/esd_max Maximum shift/esd ratio after final refinement cycle
For completeness, the following data names referring to 'esd' should also be
replaced:
_refine_ls_shift/esd_mean
Lastly, Syd proposes the following new data name to incorporate the
Nomenclature Commission's preference for the phrase "atomic displacement
parameter" rather than "thermal motion" or "thermal displacement":
_atom_site_adp_type Atom displacement parameter type (new)
_atom_site_thermal_displace_type Atom displacement type
D54.3 wR and S for refined reflections
--------------------------------------
After strong representation from authors of refinement packages, Syd has
proposed in his draft NfA the following additional data names, referring to
the wR and S values calculated from all the reflections used in the
refinement (i.e. _refine_ls_number_reflns), rather than those calculated
from "all reflections".
_refine_ls_wR_factor_ref R factor of Coeff for refined reflections (new)
_refine_ls_goodness_of_fit_ref Goodness of fit S for refined reflections (new)
D54.4 Area-detector definitions
-------------------------------
A number of participants (including Syd, David and Bill Clegg) are working
on definitions for area detector work (with some input from Andy Hammersley,
I believe). A complete set of definitions is not imminent, but Syd wishes to
include the following items for the purposes of assessing the quality of an
area-detector result set reported in Acta C:
_diffrn_detector_area_resol_mean
Resolution of area detector, in pixels/mm (new)
_diffrn_measured_fraction_theta_max
Fraction of unique refl'ns measured to q max (new)
_diffrn_measured_fraction_theta_full
Fraction of unique refl'ns measured to q full (new)
_diffrn_reflns_theta_full
\q to which available reflections are "complete" (new)
Noting Syd's wish to have a reliable set of Notes for Authors (and therefore
a reasonably stable documentation of these new data names in the form of a
beta dictionary), I would like to hear any objections to these proposals
in the near future.
Regards
Brian
- Prev by Date: (53) Observer from Nomenclature Commission
- Next by Date: (55) Call for approval: pdCIF, mmCIF
- Index(es):