Full Year 2012
----------------------
>> 13 December 2012: Simulated Annealing using MAGDOMT
----------------------
- The possibility to use the optimization of magnetic domains occupations, of type described by MAGDOMT
keyword (see note of 15 March 2009), by simulated annealing has been implemented in the current version
of FullProf.
----------------------
>> 27 November 2012: New version of BasIreps
----------------------
- There is a new version of BasIreps allowing reading user-given irreps instead of generating
them automatically using the Zak's procedure. This works only in the console application. The
GUI cannot handle directly this option.
For using this option a new keyword is read in the input file my_file.smb. The keyword is
LITTLE_GROUP and it is followed by the number of symmetry operators of Gk, the number or
irreps and the dimensions of each irrep. A negative value for a dimension means that the
irrep is complex. This keyword is immediately followed by the list of symmetry operators
and the matrices of the irreps. The template of this part of the input file is the following
(we assume that the last representation is complex)
LITTLE_GROUP Nsym_Gk N_irreps Dim_Irrep_1 Dim_irrep_2 .... -Dim_irrep_N_irreps
GK_SYMM1 x,y,z
IRREPR#1 ((GammaR1_1(i,j), j=1,Dim_Irrep1),i=1,Dim_Irrep1)
.......
IRREPR#N_irreps ((GammaRN_irreps_1(i,j), j=1,Dim_Irrep_N_irreps),i=1,Dim_Irrep_N_irreps)
IRREPI#N_irreps ((GammaIN_irreps_1(i,j), j=1,Dim_Irrep_N_irreps),i=1,Dim_Irrep_N_irreps)
GK_SYMM2 -x,y,z+1/2
IRREPR#1 ((GammaR1_2(i,j), j=1,Dim_Irrep1),i=1,Dim_Irrep1)
.......
IRREPR#N_irreps ((GammaRN_irreps_2(i,j), j=1,Dim_Irrep_N_irreps),i=1,Dim_Irrep_N_irreps)
IRREPI#N_irreps ((GammaIN_irreps_2(i,j), j=1,Dim_Irrep_N_irreps),i=1,Dim_Irrep_N_irreps)
.....
The operators and numeric values of the matrices should start after two spaces after the end of
the corresponding keywords. The matrices GammaR correspond to the module of the (in general, complex)
representation and the matrices GammaI is formed with the phases (in fraction of 2pi) of the (complex)
full matrices. An example for the space group R3c and k=(0,0,0) with real matrices is given below
------------ Start of the file BasIreps_Example1.smb ------
TITLE Example to be used with BasIreps (user-given Irreps)
SPGR 161
LITTLE_GROUP 6 3 1 1 2
GK_SYMM1 x,y,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 1 0 0 1
GK_SYMM2 -y,x-y,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 -0.5 -0.8660254 0.8660254 -0.5
GK_SYMM3 -x+y,-x,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 -0.5 0.8660254 -0.8660254 -0.5
GK_SYMM4 -y,-x,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 -0.5 0.8660254 0.8660254 0.5
GK_SYMM5 -x+y,y,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 -0.5 -0.8660254 -0.8660254 0.5
GK_SYMM6 x,x-y,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 0 1 1 0
KVEC 0.0000 0.0000 0.0000 G
BASIR AXIAL CEL
ATOM CO CO 0.15037 0.28771 0.33333333
------------- End of the file BasIreps_Example1.smb ------
The same example using an equivalent complex representation for the Irrep#3 is given below
------------ Start of the file BasIreps_Example2.smb ------
TITLE Example to be used with BasIreps (user-given Irreps, complex representation Irrep#3)
SPGR 161
LITTLE_GROUP 6 3 1 1 -2
GK_SYMM1 x,y,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 1 0 0 1
IRREPI#3 0 0 0 0
GK_SYMM2 -y,x-y,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 1 0 0 1
IRREPI#3 0.3333333 0 0 0.666666667
GK_SYMM3 -x+y,-x,z
IRREPR#1 1
IRREPR#2 1
IRREPR#3 1 0 0 1
IRREPI#3 0.666666667 0 0 0.3333333
GK_SYMM4 -y,-x,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 0 1 1 0
IRREPI#3 0 0 0 0
GK_SYMM5 -x+y,y,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 0 1 1 0
IRREPI#3 0 0.666666667 0.3333333 0
GK_SYMM6 x,x-y,z+1/2
IRREPR#1 1
IRREPR#2 -1
IRREPR#3 0 1 1 0
IRREPI#3 0 0.3333333 0.666666667 0
KVEC 0.0000 0.0000 0.0000 G
BASIR AXIAL CEL
ATOM CO CO 0.15037 0.28771 0.33333333
------------- End of the file BasIreps_Example2.smb ------
For running the above examples one has to type in a console:
> basireps BasIreps_Example1.smb
Or
> basireps BasIreps_Example2.smb
----------------------
>> 20 November 2012: Small changes in WinPLOTR-2006 and FullProf
----------------------
- Changes in WinPLOTR-2006: New facilities for reading 2D detectors at ILL, changes
in the format of the ILL-database (e.g. new D1B format is handled).
- The output file CFML_Restraints.tpcr of FullProf has now appended the number of
the phase: CFML_Restraints_k.tpcr for phase number k.
- A bug concerning flipping ratio refinements (using single pattern format) was introduced in
the previous version so that FullProf was trying to read a *.dat file instead of the usual
*.int file. This has been now corrected
----------------------
>> 2 September 2012: Changes in WinPLOTR-2006, Toolbar, GFourier. New version of WinPLOTR for Windows
----------------------
- Changes in WinPLOTR-2006 and Toolbar: The extension .xy or .xys are automatically
recognised by WinPLOTR-2006. An additional button for calculating the multipliers
of Y-axis to make a simultaneous normalisation of the loaded curves.
Remember that for saving files in column form, one has to have only a single pattern
visible in the screen (use the menu Profile > Show for changing this).
The output file in format .xys contains now information about the "scattering variable"
that now may be also "Energy", "Lambda", "Time", "Frequency" and "Temperature".
This allows to load data in which we store other things than "diffraction patterns".
When ones load this kind of data, some menu items are desactivated (in particular the
change of X-space) and are not available.
- One can now make a projection of the density separating the negative values from the
positive ones in GFourier.
- A new version of WinPLOTR for Windows is distributed. See winplotr_news.htm for details.
The setting file winplotr.set contains by default the text editor "notepad++" and Mozilla
Firefox as default browser. As usual the user may have to change these settings.
----------------------
>> 24 May 2012: Addition of a new format in DataRed. Correction of bugs.
---------------------- New output files for SuperFlip and EDMA
- A new format for reading raw integrated intensities in DataRed for preparing
single crystal input files for FullProf has been introduced after the suggestion
of Juerg Schefer (PSI) that is appropriate for incommensurate structures.
The data file contains the following items corresponding to the Fortran format
(i6,3f10.4,2f12.2,4f8.2):
Numor, Real hkls, Intensity, Sigma, 2Theta, Omega, Chi, Phi
To acces this format the user should write in the *.red file the instruction:
HKL_F 9
- By putting Fou=7 in the PCR file, FullProf produces files for the programs
SuperFlip and EDMA (http://superflip.fzu.cz). The generated files corresponding
to the PCR file "my_code.pcr" are the following:
my_code.spf ---- Contains the input parameters for SuperFlip
my_code.fou ---- Contains the reflection file in SHELX format as extracted
from a Le Bail fit.
edma.input ---- Input file for the program EDMA
- Small bugs have been corrected in GFourier (Linux, Mac).
----------------------
>> 14 April 2012: Correction of bugs. Simulated annealing with
---------------------- flipping ratios and using chi2 as cost function
- Several bugs have been corrected in the last version of the FullProf Suite:
. The generated file for DICVOL by WinPLOTR was not properly read by DICVOL.
. The binary file generated by FullProf for size and strain was not read by
GFourier.
. In some cases CIF files were still in error.
- Simulated annealing can be now be done properly with Flipping Ratio data.
- It is possible to use Chi-squared as cost fonction instead of the conventional
sum of absolute values of residuals divided by the sum of observations for
integrated intensity and profile data. The calculation of Chi-squared can be
done using four weighting schemes as described in the note of 22 March 2012.
For selected profile points, iwgt=1 means identical weight (1/mean_variance)
for all points instead of calculated variance. The iwgt=3 is restricted to
e=0.0 and q=1.0.
Four using Chi2 as cost function one has to put a negative sign to the
"Accept" value appearing in the list of parameters for tunning the simulated
annealing job.
----------------------
>> 22 March 2012: New version of FullProf. Weighting schemes. Polarised neutrons
---------------------- New keywords in COMMANDS lines.
- The current version of FullProf has been updated to:
**********************************************************'
** PROGRAM FullProf.2k (Version 5.30 - Mar2012-ILL JRC) **'
**********************************************************'
- A bug in the flipping ratio option when extinction was used has been corrected.
We, Laurent Chapon, Eric Ressouche and JRC, are re-testing the whole set of flipping ratio
options in order to improve it.
- Small bugs in the output CIF files have been corrected. No space between two Uaniso values
were appearing some times when some values were negative. Now, for multi-pattern jobs a
distinct CIF file is output for each pattern, the phases contributing to different patterns
have the same description in each individual CIF file.
- Two new keywords can now be used in COMMANDS section.
FIX_SPC spc1 spc2 spc3 ... spcN
SAME_Biso spc1 val1 spc2 val2 ... spcN valN
Where spc"i" is the symbol used for the scattering power of the atom (normally the chemical
element symbol). the command FIX_SPC means that all free parameters of a particular chemical
species are fixed. For instance, if one wants to fix all hydrogen atoms of a structure the
command "FIX_SPC H" will fix all coordinates, occupation and displacement parameters of H-atoms.
The values val"i" correspond to the value of the isotropic temperature factor to be
assigned commonly to all atoms of species spc"i". This value should be initially be given
as negative. The program will change that to positive and only when the negative value
appears in the PCR file the Biso already existing in the file are replaced by those
given in the SAME_Biso instruction.
Examples:
....
COMMANDS
.....
FIX O H
SAME_Biso O -0.93 F -0.98
.....
END_COMMANDS
- A weighting scheme similar to that used by SHELX has been implemented in FullProf.
To access this option one has to make iwgt=3 and then provide the coefficients for
the weighting scheme in the line below. For the moment the coefficients are fixed,
but it is intended to implement a refinement of some of their values.
The available weighting schemes in FullProf are summarised as follows
iwgt=0, w=1/variance(Gobs)[single crystal] or w=1/variance(yobs) [powders]
iwgt=1, w=1/variance(ycal)[powders], the variance is calculated at each cycle (Maximum Likelyhood)
iwgt=1, w=1/Average-variance[single crystals], identical weights for all reflections
iwgt=2, as iwgt=3 with default values (see below)
iwgt=3, SHELX-like weighting scheme
For single crystal work (including flipping ratio option) the expression of the weights when
iwgt=3 is:
w = q / [ variance(Gobs) + (a*P)^2 + b*P + d + e*sin(theta)/Lambda ]'
with P = f*max(0,Gobs)+(1-f)*Gcal
q=exp[c*(sin(theta)/lambda)^2] if c is positive
q=1-exp[c*(sin(theta)/lambda)^2] if c is negative
Where normally Gobs=Fo^2 and Gcal=Fc^2, but it may be the sum of contributions to an observation
due to domains or twins, or flipping ratios (Gobs=Robs, Gcal=Rcal)
Even if it has not been tested, we have implemented a weighting scheme using q=1 (c=0),
and e=0 for powders. In this case Gobs=yobs and Gcal=ycal. In both cases the set of six
coefficients a,b,c,d,e,f have to be provided in a single line starting with the keyword WGHT.
If after reading iwgt=3, no WGHT keyword line is given, the program assumes the default
values: a=0.2,b=0.0,c=0.0,d=0.0,e=0.0, f=1/3. These default values are also accessible by
simply using iwgt=2 (this has changed with respect to the previous iwgt=2 option).
Examples of WGHT keywords when iwgt=3 (in the future the change of "fixed" to "refine"
will instruct the program to optimise some of the coefficients).
For a single pattern job:
........
!Job Npr Nph Nba Nex Nsc Nor Dum Iwg Ilo Ias Res Ste Nre Cry Uni Cor Opt Aut
1 0 1 0 0 0 0 0 3 0 0 1 0 0 1 0 0 0 1
!
WGHT 0.1810 0.0120 0.0000 0.0000 0.0000 0.3333333 fixed
.....
For a multi-pattern job:
.......
!Job Npr Nba Nex Nsc Nor Iwg Ilo Res Ste Uni Cor Anm Int
0 0 0 0 0 0 3 0 0 0 0 0 0 1 !-> Patt#: 1
0 0 0 0 0 0 3 0 0 0 0 0 0 1 !-> Patt#: 2
!
WGHT 0.2110 0.0120 0.0000 0.0000 0.0000 0.3333333 fixed
WGHT 0.2000 0.0000 0.0000 0.0000 0.0000 0.3333333 fixed
!File names of data(patterns) files
.....
----------------------
>> 12 February 2012: Fourier and GFourier changes
----------------------
- By default no calculation of bond angles is done in Fourier program. The directive "List dist"
has been changed. If a third value is provided, only in this case the calculation of angles
is performed. This change has a side effect in the binary file containing the density. The
old binary files produced with Fourier/GFourier cannot be read with the present version of
the program. A new binary file should be re-generated by running the current version of the
program.
- Magnetic structure factors with sign coming from CCSL program "Sorgam" can now be read if
in the Fourier map type the keyword "msf" (without quotes) is provided.
An example calculating the spin density from a msf-file should contains two lines like:
Fourier fobs msf
File fou my_msf_file.msf
- The program Fourier can generate a binary file containing the density coming from a Maximum
Entropy calculation. Instead of calculating the Fourier map from structure factor data, the
program reads a maximum entropy map (mem), calculated using CCSL, and merges the data together
with the information given in the input file (.inp) to produce a binary file that can be read
by GFourier. For using this option the Fourier keyword should be supplemented as in the
following example:
Fourier fobs mem my_MaxEnt_file.mem
- The manual of Fourier/GFourier has been updated
----------------------
>> 8 February 2012: Problem of WinPLOTR-2006 in Windows 7 solved
----------------------
- The strange behaviour of WinPLOTR-2006 under Windows 7 has been solved. The problem
was due to the use of a big stack that apparently, in Windows 7, has a side effect on
the opening of a windows for selecting files.