CVS log for ccp4/src/mtz2various.f
	Help

Request diff between arbitrary revisions / Display revisions graphically

Add in CCP4_PROG_VERSION.
Others: please update this if mtz2various output is changed!

[#4581] increase output precision for mmCIF format

merge in 6.1.1 to 6.1.2 changes from the series-6_1 branch

bug 3963, cns and xplor formats can output data that does not contain ff, etc

Fix spacegroup information written to cif file, see #3928

Fix spacegroup information written to cif file, see #3928

Correction to fix 1.124, second cif line wasn't starting in the
right place, see bug #1557 and #2874

Correction to fix 1.124, second cif line wasn't starting in the
right place, see bug #1557 and #2874

Ok, this one works. Had to make a change I suggested in a comment elsewhere.

Thing is, I am not sure that this will not break any other formats. Logically
it should not, but I have no way of knowing. This will now produce I- when
I+ is missing

Fix for # 3162 I think - now correctly outputs statistics for I+/I- input.
Also added header update blaming me for these changes.

What did I change that I am worrying about?

if (I+ and I- and scalepack) cif_missing = False

Couldn't work out if this will have any nasty side-effects.

FIXES towards bug # 3162 as follows:

If ISCAL set sigma I's to -1.0 by default => if they are not measured then
the indication should be correct

Symmetrise - check that - reflections are considered on a par with + -
until now they were not. Have only implemented this for scalepack format.
Added fixme's elsewhere.

If not cif and I+ or I- then set Imean = 0.0: this was done already for I+
but not I-

Added a couple of FIXME statements to teh code where I want to make
changes towards 3162. Committing at this stage helps to roll back in the
event that it all goes wrong...

Catch invalid OUTPUT types

Bug #2195: patch to correct the CIF date format, which should be
"year-month-date" (Takaaki Fukami); same as revision 1.131 in main trunk.

Bug #2195: patch to correct the CIF date format, which should be
"year-month-date" (Takaaki Fukami).

Changes for SHELX output:
1. Don't output foo.ins stub at the top of HKLOUT
2. Don't output FREE at end of line
3. Apply automatic SCAL factor to I/SIGI output as well
4. Clarify some stdout

Copy revision 1.124.2.2 from patch branch

MAIN option now can output FREE column

MAIN option now can output FREE column

Array bounds check

Added in changes from main trunk for patch release: r1.25 (limits the
number of "no output" warnings for empty reflections); r1.26 (beautify
stdout); r1.27 (fix incorrect flagging of "missing" reflections - bug
#1658).

Output CIF: reflections should not be flagged as "missing" i.e. as "x" if
at least one data item is being output with real data (bug #1658).

beautified the standard output

Set a maximum to the number of "No output for this reflection" warnings
that are issued. Maximum is currently 10, see parameter MAX_WARNINGS

Bug 1557: fix problem that CIF output may break long lines in the middle
of numeric fields.

Fix to CIF output: if the output "line" extends beyond 2 lines (i.e. 160
characters) then keep adding extra lines until everything is output.

Updated CIF output to use tokens from the PDB exchange dictionary for
anomalous data and HL coefficients.

update licence banners

Change the "g" format specifiers for CIF output to "f", to prevent the
writing of data values in an invalid scientific notation (e.g. 0.1111E+5)
(bug #1195).

Allow output of FP/DP and F(+)/F(-) at the same time for CIF (bug #1191).

Allow output of FP and F(+)/F(-) to a single CIF (bug #1191).

Allow output of I and I(+)/I(-) to a single CIF (bug #1184).

Fixed poor test for I or I(+) being set to MNF, which resulted in mean
intensities being set to zero when written to CIF (bug #1179).

Made modification to apply squaring and scaling when user output
mode is selected.

From EJD: some tidying, and if input for SHELX output is F(+)/F(-) then
performing squaring automatically to produce pseudo-intensities.

Trap for invalid OUTPUT option.
Was spotted for SHELX77 which appeared to be valid option,
but is generic problem

NTOK needs to be reset for 2nd call to parser (for LABIN)

NTOK needs to be reset for 2nd call to parser (for LABIN)

EJD: stop fatal error if converting to scalepack and chosen dataset doesn't
have I(+), I(-).

For mmCIF output, increase precision of cell dimensions to 4d.p.
as in MTZ header, and increase precision of HL coeffs to 2d.p.

tidied up shelx handling for SHELXD -EJD, bug 333

add new licence banners

changes from Clemens

don't do check on output for USER option, since doesn't cover
use of DUM/IDUM labels

mmCIF option should output all reflections using '?' for MNF
This was how it was originally but got broken in 1999

mmCIF option should output all reflections using '?' for MNF
This was how it was originally but got broken in 1999

Forgot to set jfom=1 for cif output

OUTPUT SHELXD option from EJD

incorrect loop parameter

1st attempt at extended OUTPUT CIF option with extra columns
requested by EBI. Needs more testing!
Now there are explicit anomalous columns, so split into hkl/-h-k-l
remove for ICIF=1
There was also a bug when writing more than 1 cif line per reflection.

Update for crystal level (actually completely neutral)

CV - mostly mcols to 500

CV- mostly mcols to 500

Change CIF formats to ensure spaces between items.
Add COMPLEX field to OUTPUT MAIN.

Added OUTPUT MAIN option.
Also bug fix from Kim (re: sysab for OUTPUT CIF)

Changes to allow mmCIF files for EM (FC & PHIC or FP & PHIB).
Undone part of last change - speedup means no reflections output!

a user check and a speedup from ejd

Split IF in case ADATA(15) is NaN

Fix for problem with DUM9 and DUM10 labels being unavailable.

EJD:There were inconsistencies in using anom diffs.

Allow label I with CIF.
Correct treatment of centrics with CIF

For SHELX output it squares F+ - F- but not DANO, from Eleanor

Moved CIF header writing to wcifhr
Write wavelength if available from MTZ header

Place DATA statements correctly.
Add in _refln.wavelength_id, _refln.crystal_id, _refln.scale_group_code
  and parents.

Sean McSweeney: add ANOMalous= line to X-plor/CNS line
Tassos: tidy F+ -> F(+) in log output
Me: MTZ2VSCR opened as SCRATCH

ejd being tidy

A 'buglet' from ejd

Minor changes:
1. Fixed EXCLUDE DIFF option for SHELX
2. Added counter for reflections rejected by EXCLUDE k/word
3. Strengthened warning about Fs converted to Is for SHELX
4. Added summary tags for information about output format.

MDW: fix for XPLOR format problem.

Changes for CIF output:
FPART_BULK_S and PHIPART_BULK_S give extra loop columns for bulk
solvent correction.
_refln.F_meas / _refln.F_meas_sigma -> _refln.F_meas_au / _refln.F_meas_sigma_au
since these are more correct, and used by SFCHECK

Fixed writing of reflections with MNFs when keyword MISS specified.
For output CIF, these are flagged with _refln.status 'x'

Initialise ICNS. Interesting effects with g77 without it!

Minor error picked up by linux compilation.

More changes from EJD.

Changes from EJD - minor fixes for XPLOR and TNT output.

EJD: minor changes for TNT format.

Minor addition to previous changes from EJD.

Changes from Eleanor: added OUTPUT CNS option to write out CNS files.
PJB: added comments and fixed DUM..IDUM labels for OUTPUT USER option.

Fix from Laurent Maveyraud: increased number of characters allowed for
printing reflection indices in XPLOR format.

Fixed problem with DUM, IDUM labels under OUTPUT USER option, where data
was incorrectly written as zero in these columns in the output file.

Changed to new input column types for I+/- and F+/-

Attempt to restore fix from A. Leslie for XPLOR format.

Call to MSYMLB2 so that writes out correct format
spacegroup for OUTPUT CIF option.

Initialisations needed to run on g77

Fixed format statement to compile under HP-UX.

Fixed problem with OUTPUT USER printing blank lines

New version from EJD - changes to output anomalous data.

LROPEN was called with IPRINT of type LOGICAL instead of INTEGER

A y2k compatability problem

Changes from A leslie

Declarations moved above DATA for linux

Slight cock-up in last change. Free-R column now output if
FREE assigned but not used for INCLUDE | EXCLUDE FREER options.

If FREE column specified, OUTPUT TNT now outputs it with Xplor
convention. This is for Pannu + Reed version of TNT.

PRE: new option INCLUDE FREER to select just freer set

SIGMA output instead of SIGM for XPLOR.

Option to output CIF file has been included. However, can only output
FP, SIGFP I+ and I-. Also, added utility in USER mode to output integer and
reals.

Added IDUM columns and FREEVAL can accept any freeR set.

.

tidying + files CLOSED.

keyword NAN -> MISS, and argument added to allow
user to specify representation of MNF in output.

.

missing number stuff improved. Also NAN keyword will include reflections with
NaNs as zero.

EJD: allow negative I values to be written.

OUTPUT USER  can now generate DP and SIGDP. The ordering of the columns echos
the LABIN order for 'USER'. Automatic skipping of records with SIGFP=0/NaN and
SIGFPH=0/NaN is removed. All data with Magic number values are set to zero for
output file.

FREE column type should be I.

Expunge RESTRAIN i/p now restrain takes MTZ.  This was actually never
necessary since restrain can specify i/p format...

Avoid NaN lossage.

EJD: Undocumented FREEVAL keyword.  Missing data checks.
Fix free R o/p for x-plor.

check for OUTPUT given

check that FP label assigned

Tidying up.

typo

fix ccprcs arg

XPLOR output for anomalous data is now OK. Removed sigma cutoff on DP.

cosmetics re. USER o/p

DP column type was wrong

CTYP for FREE was wrong

EJD: append FREE to end of SHELX o/p for reflexions with free R flag

Mainly fix rejection with missing FPH and counting of total reflctions

TNT changes had got lost

shelx o/p was zapped by last edit

Add USER OUTPUT type.

ISYM program label wasnt typed

PRE: added TNT output option, & EXCLUDE FREER option.

EJD: read Is ready for SHELX

PRE: added options EXCLUDE FPMAX FPHMAX

.

change FRRERFLAG label to FREE, consistent with fft

integer TEST for X-PLOR

now compilable

add RESTRAIN o/p type

AGWL: Further corrections. For any phase columns
present must negate the phase when generating the Friedel
pair. This applies to solvent correction term and PHIB term.

PRE: correction to change of 17/1/94, forgot to dived
Danom by 2 before adding or subtracting to Fmean

previous change also corrected bug which could mangle indices

PRE: added option to read columns DP, SIGDP, ISYM for
anomalous differences, to write out hkl & -h-k-l if Danom present.

EJD: didnt reject reflns with SIG = 0.0

version pulled straight from york

EJD: set flag = 1 for free R included reflections

Fixes and tidying from Peter Brick and fix from PRE (not sure what for)

EJD: new FSQUARED, SCALE options

EJD+DL now o/p xplor free R flag

.

XPLOR wants PHAS not PHI

fix date printing

(PRE) mtz2various changed so that END command is not
compulsory (**ALL** programs should recognize EOF on control
input!!). Also changed so that output file ALWAYS goes to HKLOUT

Previous change should have noted:
From Jim Raftery:
does not now output EXCLUDEd reflexions; the different formats
are treated as similarly as possible; FPH being assigned will
reult in |FPH-FP| being output.

.

mods made by jim raftery to an old version

copyright

LMB: initialize IFAIL before calling CCPOPN
(how did this ever work?). Also changed CCPOPN to CCPDPN

standardisation

typo?

change ccpvrs to ccprcs

bad ccperr call

Roughly correposnds to first `official' release of MTZ version for un*x
and VMS around 14--15/5/92.

Initial revision