| People | |
| Software | |
| XDS |
| XDS (X-ray Data Streamer) |
XDS is program used for automatic processing of rotation diffraction data from crystals of initially unknown symmetry and cell constants.
To use XDS on the DEC-ALPHAs you need to add the following lines to your .login file:
alias xds /joule2/programs/xds/xds alias xscale /joule2/programs/xds/xscale
XDS needs one file called XDS.DATA that contains all the information to run the program. A typical XDS file to process RAXIS images looks like this:
IDXREF RAXIS QX=0.1017 QY=0.105 a1gil1???.osc RAXIS DIRECT NONE 1.5418 0 0 1 0.444 0.0 1.0 0.0 20 20 0.0 4.00 0 1 0 -100.0 952.0 954.0 1 0 0 0 -1 0 0.001 0 21.0 21.00 27.00 95.0 103.0 119.5 0 0 0.5 0.9 10
C***********************************************************************
C *
C *
C XX XX DDDDDD SSSSSS *
C XX XX DD DD SS *
C XX XX DD DD SS *
C XXX DD DD SSSSSS *
C XX XX DD DD SS *
C XX XX DD DD SS *
C XX XX DDDDDD SSSSSS *
C *
C *
C***********************************************************************
C *
C X-RAY DATA STREAMER FOR THE XENTRONICS AREA DETECTOR, *
C THE MAR-RESEARCH AND R-AXISII IMAGING PLATE SYSTEMS *
C VERSION August-1994 *
C *
C WOLFGANG KABSCH *
C MAX-PLANCK-INSTITUTE FOR MEDICAL RESEARCH *
C DEPARTMENT OF BIOPHYSICS *
C JAHNSTRASSE 29 *
C 69028 HEIDELBERG *
C GERMANY *
C *
C e-mail: kabsch@mpimf-heidelberg.mpg.de *
C *
C***********************************************************************
C *
C XDS is based on the theory described in the references: *
C *
C 1)"Automatic processing of rotation diffraction data from crystals *
C of initially unknown symmetry and cell constants" *
C Kabsch,W. (1993),J.Appl.Cryst.26,795-800. *
C 2)"EVALUATION OF SINGLE CRYSTAL X-RAY DIFFRACTION DATA FROM A *
C POSITION SENSITIVE DETECTOR" *
C Kabsch,W. (1988),J.Appl.Cryst.21,916-924. *
C 3)"AUTOMATIC INDEXING OF ROTATION DIFFRACTION PATTERNS" *
C Kabsch,W. (1988),J.Appl.Cryst.21,67-71. *
C *
C Instrument control and the production of data frames for the *
C Nicolet/Siemens detector are based on software described in: *
C *
C 4)"A SYSTEM FOR COLLECTION AND ON-LINE INTEGRATION OF X-RAY *
C DIFFRACTION DATA FROM A MULTIWIRE AREA DETECTOR" *
C Blum,M., Metcalf,P., Harrison,S.C. and Wiley,D.C. (1987), *
C J.Appl.Cryst.20,235-242. *
C *
C***********************************************************************
C****************** HISTORY OF PROGRAM CHANGES *************************
C***********************************************************************
C *
C ORIGINAL VERSION FOR THE NICOLET/SIEMENS DETECTOR 12-1988 *
C *
C GLOBAL PARAMETER REFINEMENT ("GLOREF"-STEP) ADDED 6-1990 *
C *
C CHANGE OF "GETFRM" (CONTRIBUTED BY GERSON COHEN, NIH) 5-1991 *
C FOR PROCESSING OF DATA COLLECTED AT THE ARGONNE NATIONAL *
C LAB. FASTER VERSION OF THE "CORRECT" SUBROUTINE. *
C *
C MAJOR MODIFICATION OF XDS FOR THE PROCESSING OF DATA 10-1991 *
C FROM CRYSTALS OF UNKNOWN SPACE GROUP SYMMETRY AND CELL *
C CONSTANTS. *
C *
C BUG CORRECTIONS AND MINOR CHANGES 2-1992 *
C *
C ADAPTATION TO THE MAR RESEARCH IMAGING PLATE 8-1992 *
C *
C PRE-REFINEMENT ADDED TO "IDXREF" 11-1992 *
C *
C HANDLING OF HARVARD 16-BIT FRAME FORMAT ADDED. 8-1993 *
C XDS NOW ACCEPTS UNIX COMPRESSED DATA FRAME FILES. *
C COMPRESS-COMMAND. THE MANUAL INCLUDES A REPORT OF A *
C CASE OF OBTAINING THE WRONG ENANTIOMORPH BY DATA *
C PROCESSING WITH XDS (CONTRIBUTED BY JANET SMITH, PURDUE) *
C UTILITY FUNCTION "DIRB" ADDED. IT PROVIDES A TOOL FOR *
C ESTABLISHING THE MAPPING BETWEEN PIXEL-SPACE AND THE LAB *
C COORDINATE SYSTEM. *
C *
C ALL RESTRICTIONS FOR THE MAXIMUM NUMBER OF REFLECTIONS 1-1994 *
C CLOSE TO THE EWALD SPHERE HAVE BEEN ELIMINATED. *
C THE MINIMUM NUMBER OF CONSECUTIVE DATA FRAMES REQUIRED *
C BY "XDS" HAS BEEN REDUCED TO 1. *
C *
C ADAPTATION TO THE R-AXIS IMAGING PLATE DETECTOR 8-1994 *
C *
C***********************************************************************
C****************** INSTALLATION OF THE PROGRAM ************************
C***********************************************************************
C *
C THE PROGRAM IS WRITTEN IN FORTRAN 77 WITH FEW NON-STANDARD *
C EXTENSIONS OF THE LANGUAGE DEFINITION. THESE ARE: *
C *
C NON-STANDARD DATA TYPES *
C *
C It is assumed in XDS that: *
C A CHARACTER*1 VARIABLE IS REPRESENTED BY 8 BITS=1 BYTES. *
C A INTEGER*2 VARIABLE IS REPRESENTED BY 16 BITS=2 BYTES. *
C A INTEGER*4 VARIABLE IS REPRESENTED BY 32 BITS=4 BYTES. *
C A INTEGER VARIABLE IS REPRESENTED BY 32 BITS=4 BYTES. *
C A REAL VARIABLE IS REPRESENTED BY 32 BITS=4 BYTES. *
C A REAL*4 VARIABLE IS REPRESENTED BY 32 BITS=4 BYTES. *
C A REAL*8 VARIABLE IS REPRESENTED BY 64 BITS=8 BYTES. *
C *
C *
C NON-STANDARD FUNCTIONS AND SUBROUTINES *
C *
C XDS uses the non-standard functions "IAND", "IOR" and the subroutine *
C "DATE". With few exceptions these are available on most computers *
C and no modifications are required. Known exceptions are PCS/CADMUS *
C and IBM risk 6000 computers. *
C PCS/CADMUS: Uncomment the code provided at the end of this file *
C (xds.f) for "IAND", "IOR". (Use your editor to search *
C for "CADMUS".) *
C IBM RISC 6000: Uncomment the code for subroutine "DATE" provided *
C at the end of this file (xds.f). (Use your editor *
C to search for "DATE".) *
C *
C SYSTEM CALLS *
C *
C All system calls are issued in subroutine OPFRAM located near the *
C end of this file (xds.f). Appropriate calls are described for the *
C UNIX and VAX-VMS operating systems. Use your editor to search for *
C "SYSTEM(" to check and possibly modify the system calls according *
C to the comments near these places. *
C THE PROGRAM CODE IS CONTAINED ON TWO FILES: *
C *
C - THE FILE "xdslib.f" contains a large number of hopefully device *
C independent subroutines needed by the main program. It is unlikely *
C that you have to modify anything on this file. *
C *
C - THE FILE "xds.f" (the one you read now) contains the manual, main *
C program and a small number of probably machine dependent routines. *
C All constants used by XDS are named by symbols and assigned values *
C at a unique place in the main program. Some of these constants *
C control the size of arrays used by the subroutines. These arrays *
C are allocated here in the main program (file xds.f) to simplify a *
C change of their sizes without the necessity to recompile all the *
C subroutines (file xdslib.f). To make a new version of XDS tailored *
C to a specific task (like processing data taken from crystals with *
C a very large unit cell) thus amounts to changing a few constants *
C and recompiling the main program only. *
C COMMON and EQUIVALENCE directives are used to reduce the size of the *
C virtual address space for the whole program since not all of the *
C arrays are needed at the same time in each program step. XDS is *
C organized into 7 steps - XYCORR, COLSPOT, IDXREF, COLPROF, *
C PROFIT, CORRECT, GLOREF - which are invoked consecutively by the *
C main program. There is no exchange of information between these *
C steps through shared memory. Information is exchanged by files *
C which allows to repeat selected steps in case of a system crash *
C or with changed input parameters in XDS.DATA without rerunning the *
C whole program. *
C In the following you find a description of what you have to do for *
C installing XDS at your specific environment. In addition, some *
C examples of global parameter settings are given. *
C *
C GLOBAL PARAMETER SETTINGS *
C *
C USE YOUR EDITOR TO SEARCH FOR "FRMTYP=" TO CHECK AND POSSIBLY MODIFY *
C ITS VALUE WHICH SETS THE DEFAULT FOR TYPE AND ACCESS OF DATA FRAMES. *
C Values 1,...,4 for FRMTYP are applicable to the NICOLET/SIEMENS *
C multiwire area detector, while 5 or 6 should be used for the MAR- *
C research imaging plate system, 7 or 8 for the R-AXIS. *
C FRMTYP=1 IF DATA FRAMES (ORIGINAL HARVARD FORMAT) ARE OPENED WITH *
C ACCESS='DIRECT', AS IS THE CASE FOR CONVEX, PCS, IRIS, *
C IBM-RISK 6000, DEC 3000/400 (OSF/1). *
C =2 IF DATA FRAMES (ORIGINAL HARVARD FORMAT) ARE OPENED WITH *
C ACCESS='SEQUENTIAL', AS IS THE CASE FOR A VAX (VMS). *
C =3 IF DATA FRAMES (ARGONNE FORMAT) ARE OPENED WITH ACCESS= *
C 'DIRECT'.(CONVEX,PCS,IRIS,IBM-RISK6000, DEC 3000/400) *
C =4 IF DATA FRAMES (ARGONNE FORMAT) ARE OPENED WITH ACCESS= *
C 'SEQUENTIAL', AS FOR A VAX (VMS). *
C =5 IF DATA FRAMES (MAR RESEARCH FORMAT) ARE OPENED WITH *
C ACCESS='DIRECT'. (CONVEX,PCS,IRIS,IBM-RISK6000,DEC 3000/400)*
C =6 IF DATA FRAMES (MAR RESEARCH FORMAT) ARE OPENED WITH *
C ACCESS='SEQUENTIAL'. (VAX VMS) *
C =7 IF DATA FRAMES (R-AXIS II 1900 X 1900) ARE OPENED WITH *
C ACCESS='DIRECT'. (CONVEX,PCS,IRIS,IBM-RISK6000,DEC 3000/400)*
C =8 IF DATA FRAMES (R-AXIS II 1900 X 1900) ARE OPENED WITH *
C ACCESS='SEQUENTIAL'. (VAX VMS) *
C IN "XDS.DATA" YOU MAY ALWAYS OVERRIDE THIS DEFAULT SETTING BY *
C USING OPTIONAL KEY-WORDS "HARVARD", "ARGONNE", "MAR", "DIRECT", *
C "SEQUENTIAL" FOLLOWING THE FILE NAMES OF THE FRAME CONTAINING SPATIAL*
C CORRECTION DATA AND THE NAME TEMPLATE OF FRAME FILES. *
C *
C USE YOUR EDITOR TO SEARCH FOR "A=5." TO CHECK AND POSSIBLY MODIFY *
C ITS VALUE. "A" IS THE SHORTEST DISTANCE (mm) BETWEEN HOLES IN THE *
C CALIBRATION BRASS PLATE FOR THE NICOLET/SIEMENS DETECTOR. *
C *
C Other parameter values which are important for the type of detector *
C you have, are QX,QY,NX,NY. A description of the meaning *
C of all parameters used by XDS is included in this manual (xds.f). *
C Some typical sets of parameter values are given below for the *
C SIEMENS and the MAR-research detectors. *
C *
C PARAMETER SIEMENS R-AXIS II MAR *
C----------------------------------------------------------------------*
C A 5.08 not used not used *
C DELPHI 2.5 5.0 5.0 *
C FRMTYP 1...4 7 5 or 6 *
C MAXRFL 160000 300000 300000 *
C NGRID 4 4 4 *
C NGRIDG 4 4 4 *
C NX 512 1900 1200 *
C NY 512 1900 1200 *
C QX 0.19 0.1017 0.15 *
C QY 0.19 0.105 0.15 *
C OVLOAD 65000 250000 130000
C TBLSIZ 32749 299993 299993 *
C----------------------------------------------------------------------*
C Program 16Mbyte 24Mbyte 24Mbyte *
C size *
C *
C *
C COMPILATION *
C *
C AFTER THE MODIFICATIONS AND CHECKS OF xds.f HAVE BEEN CARRIED OUT, *
C COMPILE AND LINK xds.f, xdslib.f TO GET AN EXECUTABLE PROGRAM. *
C IF AVAILABLE, IT IS A GOOD IDEA TO USE A COMPILER FLAG THAT *
C GENERATES REAL NUMBERS IN IEEE-STANDARD REPRESENTATION. THIS *
C RESULTS IN MUCH LESS TROUBLES IF YOU TRY TO USE XDS-GENERATED *
C BINARY FILES ON OTHER MACHINES. *
C *
C***********************************************************************
C************* COORDINATE SYSTEMS AND SAMPLE INPUT DATA ****************
C***********************************************************************
C *
C Among the input data the user has to provide (see description of *
C file XDS.DATA for details) are geometrical parameters describing *
C the setup of the diffraction experiment with respect to a fixed *
C laboratory coordinate system. Any convenient right-handed *
C orthonormal system may be chosen with the origin at the inter- *
C section between rotation axis, direct beam and crystal. *
C *
C Using the laboratory coordinate system as described in reference 3), *
C the y-axis points vertically down and is defined to be collinear *
C (parallel or antiparallel) with the rotation axis. The z-axis lies *
C within the plane spanned by the rotation axis and the detector *
C normal at swing angle 0, with +z pointing from the crystal towards *
C the detector. The x-axis is defined to yield an orthonormal right *
C handed laboratory coordinate system {x,y,z}. *
C A typical experimental setup is shown below: *
C *
C viewed from ABOVE: *
C *
C \ *
C \ *
C | \ *
C | \ *
C | \ detector *
C | \ *
C | \ *
C ------beam->-->-->----------O------------------- +z \ *
C | \ *
C | *
C | *
C | *
C | *
C +x *
C *
C *
C viewed from the SIDE: *
C *
C omega axis *
C 2-theta axis *
C rotation axis *
C | *
C | | *
C | | *
C | | *
C | | *
C ------beam->-->-->----------O------------------- +z | detector *
C | | *
C | | *
C | | *
C | | *
C | *
C +y *
C *
C Direct beam: *
C The positive beam direction points along +z from the source towards *
C the crystal. If the beam is aligned normal to the rotation axis and *
C the detector normal at swing angle 0, the coordinates of the incident*
C beam wavevector (see input parameter S0 in XDS.DATA) are 0 0 1. *
C *
C Rotation axis: *
C XDS requires that the frame increment (oscillation range, see input *
C parameter DPHI in XDS.DATA) be positive. Thus the coordinates of the *
C rotation axis (see input parameter ACHSE in XDS.DATA) are: *
C 0 1 0 if the crystal rotates clockwise when proceeding to the *
C next data frame and the camera is viewed from above. *
C 0 -1 0 if the crystal rotates counterclockwise. *
C *
C Detector position: *
C The swing axis coincides with the rotation axis and a positive swing *
C angle CHI (or 2-theta) corresponds to a clockwise rotation of the *
C detector when viewed from above. In the drawing, the detector is set *
C to a negative swing angle CHI. The orientation of the detector with *
C respect to this laboratory system is described by a right-handed set *
C of 3 orthonormal vectors ED(.,1), ED(.,2), ED(.,3) (see XDS.DATA). *
C Only the first two unit vectors must be specified while the third one*
C is automatically determined by XDS. The vectors ED are found from the*
C requirement that a pixel on the detector surface which is represented*
C by the array indices IX, IY in the data frame must map to the lab *
C coordinates according to the equations *
C x=QX*(IX-XORG)*ED(1,1)+QY*(IY-YORG)*ED(1,2)+F*ED(1,3) *
C y=QX*(IX-XORG)*ED(2,1)+QY*(IY-YORG)*ED(2,2)+F*ED(2,3) *
C z=QX*(IX-XORG)*ED(3,1)+QY*(IY-YORG)*ED(3,2)+F*ED(3,3). *
C Here, QX and QY are the size of a pixel (mm) along IX and IY, *
C respectively. F is the detector to *
C crystal distance (mm), and XORG, YORG are the pixel origin. Hence, *
C the correct choice of the vectors ED not only depends upon the swing *
C angle CHI but also upon the internal representation of a pixel by the*
C array indices IX, IY in each data frame. For HARVARD-style frames, as*
C interpreted by XDS, pixel coordinates IX=IY=1 correspond to the lower*
C right hand corner of the detector when viewed from the crystal and *
C the vectors ED(.,1), ED(.,2) are antiparallel to the laboratory x and*
C y axes at swing angle CHI (or 2-theta) = 0. The computed vector *
C ED(.,3) then points along +z which leads to a positive detector *
C distance F. Hence, for a swing angle CHI and HARVARD-style frames, *
C we have ED(1,1)=-cos(CHI), ED(2,1)= 0.0, ED(3,1)=sin(CHI) *
C ED(1,2)= 0.0, ED(2,2)=-1.0, ED(3,2)= 0.0 *
C and a positive detector distance F. (Recall that CHI is negative in *
C the example illustrated above.) *
C *
C======================================================================*
C SAMPLE INPUT DATA FOR THE SPECIAL CASE OF A SUPPER ROTATION CAMERA *
C......................................................................*
C XDS *
C SIEMENS /../urml/data/ADN900/xyco2303.12cm DIRECT HARVARD *
C/../urml/data/ADN900/ADN900.???? DIRECT HARVARD *
C/../urml/data/ADN.native *
C 1.5418 0.0 0.0 1.0 0.5 0.0 1.0 0.0 *
C 1 600 0.0 0.1666666 0.0 1.0 0.0 *
C 120.0 256.5 237.3 -0.996195 0.0 -0.087156 0.0 -1.0 0.0 0.001 *
C 19 133.0 56.4 110.0 90.0 90.0 90.0 *
C 30 1 0.45 0.45 60 *
C......................................................................*
C======================================================================*
C SAMPLE INPUT DATA FOR THE X100A SIEMENS DETECTOR AT EMBL HEIDELBERG *
C......................................................................*
C XDS *
C SIEMENS data$x:[tucker]040893.brs9 ARGONNE DIRECT *
C data$x:[tucker.p20]p20.???? ARGONNE DIRECT *
C NONE *
C 1.5418 0 0 1 0.444 0 1.0 0 *
C 1 360 0.0 0.1 0 -1 0 *
C 90 258 253 -0.9063077 0.0 0.42261826 0 -1 0 0.001 *
C 91 90 90 74 90 90 90 *
C 5 1 0.50 0.30 30 *
C......................................................................*
C Note that the rotation axis is 0 -1 0 because a positive rotation is *
C counterclockwise when looking down Y for the 3-axis goniostat used *
C at EMBL-Heidelberg. In the example above CHI was negative; the *
C direction cosines for the detector X-axis ED(.,1) are *
C *
C -cos(swing) 0 -sin(swing) *
C *
C======================================================================*
C - SAMPLE INPUT DATA FOR THE MAR RESEARCH IMAGING PLATE DETECTOR *
C......................................................................*
C XDS *
C MAR NPIX=1200 ROFF=0.0 TOFF=0.0 *
C data$m:[tucker.test]l20_???.image MAR DIRECT *
C NONE *
C 1.5418 0 0 1 0.444 0.0 1.0 0.0 *
C 1 42 0.0 1.00 0 1 0 *
C 90 595.0 595.0 1 0 0 0 1 0 0.001 *
C 96 79.35 79.35 38.34 90 90 90 *
C 30 1 0.5 0.4 10 *
C......................................................................*
C======================================================================*
C - SAMPLE INPUT DATA FOR THE R-AXIS II IMAGING PLATE DETECTOR *
C......................................................................*
C XDS *
C RAXIS QX=0.1017 QY=0.105 *
C ../frames/V43M1???.OSC RAXIS DIRECT *
C NONE *
C 1.5418 0 0 1 0.444 0.0 1.0 0.0 *
C 1 21 0.0 1.50 0 1 0 *
C -69.4305 951.8 952.8 1 0 0 0 -1 0 0.001 *
C 168 91.2 91.2 45.87 90.0 90.0 120.0 *
C 0 1 0.5 0.2 10 *
C......................................................................*
C======================================================================*
C *
C CHECKING THE INTERNAL REPRESENTATION OF FRAME DATA *
C A simple way of checking how a pixel on the detector surface is *
C represented by its array indices IX, IY in the data frame as inter- *
C preted by XDS consists of the following procedure: *
C Move the detector to swing angle 0. Sucessively expose the upper *
C and lower right corners of the detector surface to x-rays and save *
C each exposure in a separate data frame file. *
C Use XDS to find the pixel coordinates IX, IY of the maxima in each *
C of the 2 data frames by specifying "DIRB" in first line of XDS.DATA. *
C (Do not use anybody elses program as it might have adopted different *
C conventions!) *
C XDS will then enter interactive mode, ask for the name of the frame *
C data file and returns the pixel coordinates of the maximum. From *
C the results it is easy to figure out which way the array indices *
C IX, IY are running with respect to the laboratory coordinate system. *
C For example, at our installation the maxima were found at *
C IX=139, IY=482 in the upper right, and IX=125, IY=75 in the lower *
C right corner of the detector when viewed from the crystal. Hence, *
C in this case the IX- and IY-lines must be going antiparallel to the *
C laboratory x and y axes at swing angle 0. For a swing angle CHI *
C (remember that CHI is negative in the drawing above) this leads to *
C ED(1,1)=-cos(CHI), ED(2,1)= 0.0, ED(3,1)=sin(CHI) *
C ED(1,2)= 0.0, ED(2,2)=-1.0, ED(3,2)= 0.0 *
C The vector ED(.,3) (computed by XDS) then points mainly along +z *
C which leads to a positive detector distance F. *
C *
C INCORRECT ENANTIOMORPH (contributed by Janet Smith,Purdue University)*
C An incorrect choice of the vectors ED may well lead to incorrect *
C signs of the anomalous intensity differences as pointed out by *
C Janet Smith (Purdue University, USA). Because of its importance *
C to other users of XDS, Janet's report is included here: *
C We have encountered an enantiomorph problem in data that were *
C obtained in the following way: *
C *
C 3-axis Siemens camera, Siemens software *
C Siemens multiwire detector *
C XDS indexing and integration *
C *
C Following our reading of the XDS documentation, data in the incorrect*
C enantiomorph were produced. We have obtained the correct sign for *
C anomalous differences with the following input parameters for the *
C specified experiment (see drawing above). *
C *
C direct beam axis: 0.0 0.0 1.0 *
C rotation axis: 0.0 1.0 0.0 *
C detector vectors: -cos(CHI) 0.0 sin(CHI) *
C 0.0 -1.0 0.0 *
C detector distance: positive *
C *
C Our problem came in defining positive x and y directions on the *
C detector. The example in the XDS documentation for the "special case *
C of a 3-axis goniometer" led us in the wrong direction. It now appears*
C that the example for the "special case of a Supper rotation camera" *
C fits exactly with the situation of the three-axis camera and Harvard-*
C style frames that we use in our laboratory. *
C *
C While solving our enantiomorph problem, we tried all possible sign *
C combinations for the nonzero elements of the detector x and y vectors*
C and for the detector distance. Only two sign combinations, of 32 *
C total, could be indexed and integrated by XDS. For the experimental *
C setup illustrated above, these were: *
C *
C -cos(CHI) 0.0 sin(CHI) *
C correct enantiomorph 0.0 -1.0 0.0 *
C positive distance *
C *
C and *
C *
C -cos(CHI) 0.0 sin(CHI) *
C incorrect enantiomorph 0.0 1.0 0.0 *
C negative distance *
C *
C - MISLEADING EXAMPLE FOR THE SPECIAL CASE OF A 3-AXIS GONIOMETER *
C......................................................................*
C XDS *
C SIEMENS /../urml/data/ADN900/xyco2303.12cm DIRECT HARVARD *
C/../urml/data/ADN900/ADN900.???? DIRECT HARVARD *
C/../urml/data/ADN.native *
C 1.5418 0.0 0.0 1.0 0.5 0.0 1.0 0.0 *
C 1 600 0.0 0.1666666 0.0 1.0 0.0 *
C -120.0 256.5 237.3 -0.996195 0.0 -0.087156 0.0 1.0 0.0 0.001 *
C 19 133.0 56.4 110.0 90.0 90.0 90.0 *
C 30 1 0.45 0.45 60 *
C......................................................................*
C The direct beam direction has coordinates: 0.0 0.0 1.0 *
C The rotation axis has coordinates: 0.0 , 1.0 , 0.0 *
C The swing axis coincides with the rotation axis and a positive *
C swing angle CHI corresponds to a clockwise rotation when looking *
C in the direction of the swing axis. Be careful with the sign of *
C CHI !!! *
C Using data frames as produced by the HARVARD software (reference 4) *
C we have for a detector swing angle CHI *
C ED(1,1)=-cos(CHI) ED(2,1)= 0 ED(3,1)=-sin(CHI) *
C ED(1,2)= 0 ED(2,2)= 1 ED(3,2)= 0 *
C Since the vector ED(.,3) is pointing from the detector towards the *
C crystal we get a negative sign for "F". *
C ------------- End of misleading example ------------------ *
C***********************************************************************
C******************* HOW TO RUN THE XDS-PROGRAM ************************
C***********************************************************************
C *
C Very often, failure of the XDS-program to perform its job is caused *
C by incorrect values of the input parameters in XDS.DATA or a mis- *
C aligned experimental setup. A simple procedure to obtain sufficiently*
C correct values is described below. *
C *
C 1.) CHECK INTERNAL REPRESENTATION OF FRAME DATA *
C This check is carried out as described above under the topic *
C COORDINATE SYSTEMS AND SAMPLE INPUT DATA. This has to be done *
C only once for your installation. An incorrect choice of the *
C detector vectors ED may well lead to incorrect signs of the *
C anomalous intensity differences. *
C *
C 2.) CORRECTION VALUE FOR THE CRYSTAL TO FILM DISTANCE *
C Collect a data set of, say 50 frames from a crystal of known cell *
C parameters at swing angle 0 and process using FIXF=0 on the XDS- *
C input data cards. The correction value is the difference between *
C the refined value and the scale reading on the instrument. This *
C will be sufficiently accurate and has to be done only once for *
C the installation. *
C *
C 3.) CHECK CAMERA ALIGNMENT *
C Move the detector to swing angle 0 and check that it is o.k. Then *
C attenuate the direct beam and find its pixel positions on the *
C detector at two different crystal to detector distances, say 10 *
C and 20 cm, by using the 'DIRB' option of XDS. The easiest *
C procedure is to align the direct beam such that it intersects the *
C crystal and the rotation axis and gives the same pixel readings at *
C both 10 and 20 cm to within one pixel. The direct beam direction *
C is then normal to the detector and the detector origin is the *
C pixel reading of the direct beam position recorded earlier. *
C *
C 4.) MEASURE SPATIAL CORRECTION (NICOLET/SIEMENS detector) *
C After flat field correction mount the brass grid plate on the *
C detector face and move the detector to the distance you intend to *
C use for data collection. Put an iron x-ray source exactly (!) at the *
C place later occupied by the crystal (the origin of the laboratory *
C coordinate system). A misplaced x-ray source will result into *
C wrong correction values computed by XYCORR which might prevent a *
C proper indexing of the diffraction spots. *
C Collect the detector response for about 60min (depends on the source)*
C and save on a file with a name of your choice. Remove brass plate *
C and iron source. Don't forget to reset the high voltage of the *
C detector to CuKa-value. *
C *
C 5.) PREPARE INPUT DATA *
C To avoid confusion with other data create a new subdirectory and *
C move to it. During the run of the program a number of files will be *
C created in the current directory with predefined names. *
C Create a file with name "XDS.DATA" in the current directory. *
C This file must contain the input data specific for your data *
C collection run. The format and the values you have to provide are *
C described in detail in the section about file structures. A sample *
C XDS.DATA file is listed above. *
C *
C 6.) START DATA COLLECTION (NICOLET/SIEMENS detector) *
C Move the detector to the swing angle and the distance you want to *
C to use for data collection and mount the beam stop. *
C Start the HARVARD program to collect a sequence of consecutive *
C data frames. Start XDS-program. As data collection proceeds a *
C number of files are created which have the name extension ".LP". *
C These files report the results (and troubles) from each step. *
C If everything went well the main results you want to keep are *
C on files "XDS.HKL" (corrected intensities of all reflections) *
C and "UNIQUE.HKL" (symmetry averaged reflection intensities and *
C anomalous intensity differences). However, it is advisible to save *
C all files on streamer or tape. *
C *
C***********************************************************************
C********************** DESCRIPTION OF FILES ***************************
C***********************************************************************
C *
C DEFINITION OF FILE-TYPES USED *
C *
C TEXT = FORMATTED SEQUENTIAL HUMAN READABLE ASCII-TEXT *
C BIN = UNFORMATTED SEQUENTIAL VARIABLE LENGTH RECORD SIZE *
C DIR = UNFORMATTED DIRECT ACCESS FIXED LENGTH RECORD SIZE *
C *
C***********************************************************************
C *
C DEVICE FILE-NAME TYPE CONTENTS *
C NUMBER *
C *
C BKGPIX BKGPIX.XDS DIR BACKGROUND AT EACH DETECTOR PIXEL *
C CREATED BY: COLSPOT *
C USED BY : COLPROF *
C CDR XDS.DATA TEXT USER SUPPLIED INPUT DATA *
C CREATED BY: CRYSTALLOGRAPHER *
C USED BY : XDS,XYCORR,COLSPOT, *
C IDXREF,COLPROF,PROFIT, *
C CORRECT,GLOREF *
C LP TEXT PRINTED OUTPUT FROM EACH PROGRAM STEP *
C XYCORR.LP CREATED BY: XYCORR *
C COLSPOT.LP CREATED BY: COLSPOT *
C IDXREF.LP CREATED BY: IDXREF *
C COLPROF.LP CREATED BY: COLPROF *
C PROFIT.LP CREATED BY: PROFIT *
C CORRECT.LP CREATED BY: CORRECT *
C GLOREF.LP CREATED BY: GLOREF *
C MISFIT MISFITS TEXT LISTING OF MISFITS RECOGNIZED BY *
C COMPARISON WITH SYMMETRY EQUIVALENT *
C REFLECTIONS. *
C CREATED BY: CORRECT *
C USED BY : USER FOR INSPECTION *
C SPOT SPOT.XDS TEXT LIST OF STRONG DIFFRACTION SPOTS *
C CREATED BY: COLSPOT *
C USED BY : IDXREF *
C TAPREF ------- TEXT OPTIONAL REFERENCE DATA SET (FILE-NAME IS *
C SPECIFIED BY USER IN "XDS.DATA") *
C CREATED BY: XDS-FILE OF TYPE UNIQUE.HKL *
C FROM A PREVIOUS RUN *
C USED BY : CORRECT *
C TAPOTA UNIQUE.HKL TEXT SYMMETRY AVERAGED REFLECTION INTENSITIES *
C AND ANOMALOUS DIFFERENCES AFTER SCALING *
C CREATED BY: CORRECT *
C USED BY : CRYSTALLOGRAPHER *
C TAPOTB XDS.HKL DIR SCALED INTENSITIES OF ALL REFLECTIONS *
C WITHOUT SYMMETRY AVERAGING *
C CREATED BY: CORRECT *
C USED BY : CRYSTALLOGRAPHER *
C XFRM ------- --- CONNECTED WITH EACH DATA FRAME IN *
C SUCCESSION *
C CREATED BY: HARVARD PROGRAM *
C USED BY : XYCORR,COLSPOT,COLPROF *
C XPARM XPARM.XDS TEXT PARAMETERS DESCRIBING DIFFRACTION GEOMETRY*
C CREATED BY: IDXREF *
C USED BY : COLPROF *
C XPROF PROFIT.HKL DIR INTEGRATED INTENSITIES (UNSCALED) *
C CREATED BY: PROFIT *
C USED BY : CORRECT,GLOREF *
C XREC XREC.XDS BIN INDIVIDUAL REFLECTION PROFILES *
C CREATED BY: COLPROF *
C USED BY : PROFIT *
C XYCTBL XYCORR.TABEL DIR SPATIAL CORRECTION LOOK-UP TABLE *
C CREATED BY: XYCORR *
C USED BY : COLSPOT,IDXREF,COLPROF *
C *
C***********************************************************************
C *
C XDS.DATA (formatted sequential) *
C ======== *
C *
C This file contains the input parameters you have to provide to run *
C the XDS program.(free format) *
C *
C line # DESCRIPTION OF INPUT PARAMETERS *
C *
C 1 (JOB CARD)List of subroutine calls issued by xds. *
C The following routines may be specified in arbitrary order: *
C DIRB - XDS will read the JOB- and DEVICE CARD (the first *
C two cards in XDS.DATA), enter terminal mode and *
C ask for name, access and type of a data frame. It *
C reports the coordinates IX, IY of the pixel with *
C the largest number of counts and stops. All other *
C routine calls which may be present on this first *
C line of XDS.DATA will be ignored. If access and/ *
C or type of the data frame are omitted, defaults *
C are used according to the value of the parameter *
C "FRMTYP". A compressed data frame is recognized *
C automatically by the suffix .Z attached to the *
C regular file name. *
C XDS - complete run (normal case), which will call *
C automatically all of the routines listed below. *
C XYCORR - calculate spatial corrections *
C COLSPOT - collects strong diffraction spots from the first *
C few data frames in the range FFRAME...LFRAME. *
C IDXREF - find crystal orientation from spot locations *
C and refine all parameters. *
C COLPROF - collect 3-dimensional reflection profiles *
C PROFIT - estimate reflection intensities from profiles *
C CORRECT - correct intensities for decay, absorption and *
C detector surface sensitivity *
C GLOREF - refines all diffraction parameters thereby *
C using all observed spots from this data *
C collection run. The unit cell parameters are *
C much more reliably estimated than during the *
C profile collection step. This, however, is only *
C true if the crystal has not slipped and the *
C cell parameters have not changed during data *
C collection. *
C 2 (DEVICE CARD)The first key-word on this card designates the *
C detector used and must be either SIEMENS or MAR or RAXIS. *
C For the SIEMENS detector, the file name of the brass- *
C plate image must follow. This name must not be longer than *
C 50 characters. The name may be followed by the optional *
C key-words DIRECT or SEQUENTIAL (at most one of these two) *
C and HARVARD or ARGONNE (at most one of these two) to over- *
C ride the default setting of the global parameter "FRMTYP". *
C The key-words are recognized by their first three characters*
C The effect of the key-words is restricted to the file *
C containing spatial correction data (the brass-plate image). *
C For the MAR detector, the key-word may optionally be *
C followed by NPIX=1200 or 2000 to specify the number of *
C pixels along a scan-line for each image, and correction *
C values ROFF=radial correction TOFF=tangential correction *
C for the spiral read-out. Default settings are NPIX=1200 *
C ROFF=0.0 TOFF=0.0. *
C For the RAXIS detector, the key-word may optionally be *
C followed by NPIX=1900 or 950 to specify the number of *
C pixels along a scan-line for each image, and *
C QX=size of a pixel along a scan line [mm] *
C QY=distance between scan lines [mm]. *
C Default settings are NPIX=1900 QX=0.1017 QY=0.105 *
C 3 Name template of data frames. A sequence of 3 (MAR or RAXIS)*
C or 4 (SIEMENS) question marks within the template indicates *
C the position for the actual frame number. This number will *
C be inserted by xds as required. Intervening blanks within *
C the name template are not allowed. *
C The name template may be followed by the optional key-words *
C DIRECT or SEQUENTIAL (at most one of these two) and HARVARD,*
C ARGONNE, MAR or RAXIS (at most one of these four) to over *
C ride the default setting of the global parameter "FRMTYP". *
C The effect of the key-words is restricted to all data *
C frames with the given name template. *
C 4 File name of a reference data set used for local scaling. *
C Each record of this formatted sequential file consists of *
C h,k,l,INTENSITY,SIGMA plus optional information. *
C Only the first 5 numbers of each record will be read (free *
C format). The reflections need not be sorted. *
C If no reference data set is available, File name= NONE. *
C 5 FLAMDA- x-ray wavelength (Angstroem) *
C S0 - x,y,z components of the direct beam wavevector with *
C respect to the laboratory coordinate system. *
C The length of S0 will be set to 1/FLAMDA by XDS. *
C POL - Fraction of polarization of direct beam in a plane *
C specified by its normal 'POLPLA'. 0<=POL<=1 *
C POLPLA- x,y,z components of polarization plane normal *
C For an unpolarized beam hitting the crystal: *
C POL= 0.5 *
C POLPLA = any vector perpendicular to S0 *
C For a monochromator with incident unpolarized beam: *
C POL= [cos(2*thetaM)]**2/[1+(cos(2*thetaM))**2] *
C POLPLA = components of the monochromator *
C diffraction plane normal. *
C 6 FFRAME- The number of the first data frame collected. *
C LFRAME- The number of the last data frame collected. *
C Using the editor, you may change this number during *
C the run of subroutines COLPROF or COLSPOT. This may *
C be useful if the crystal died earlier ( or lasts *
C longer ) than expected. *
C PHI0 - Spindle position (degrees) at beginning of first *
C data frame (that is frame number 1) *
C DPHI - Oscillation range (degrees)covered by each data *
C frame. DPHI must be positive. *
C ACHSE - x,y,z components of the rotation axis with respect *
C to the laboratory system. The length of this vector *
C will be normalized by the program. *
C 7 F - Signed distance between crystal and detector (mm). *
C Note that "F" may be negative !!! *
C ORGX - Detector X-coordinate (pixels) of origin, *
C ORGY - Detector Y-coordinate (pixels) of origin. *
C The coordinates of the origin are not corrected *
C for spatial distortion !!! *
C The corrected origin is defined as that point on *
C the detector which is closest to the crystal. *
C ED - Orientation matrix of the detector with respect to *
C the laboratory system. The lab coordinates of *
C a point IX,IY (pixels) on the detector are, *
C assuming that IX,IY are already corrected for *
C spatial distortion and XORG=corrected(ORGX), *
C YORG=corrected(ORGY) : *
C x=QX*(IX-XORG)*ED(1,1)+QY*(IY-YORG)*ED(1,2)+F*ED(1,3) *
C y=QX*(IX-XORG)*ED(2,1)+QY*(IY-YORG)*ED(2,2)+F*ED(2,3) *
C z=QX*(IX-XORG)*ED(3,1)+QY*(IY-YORG)*ED(3,2)+F*ED(3,3) *
C Here, QX, QY are the lengths of a pixel (mm) along *
C IX and IY, respectively. *
C Only two orthogonal unit vectors must be given: *
C ED(1,1),ED(2,1),ED(3,1),ED(1,2),ED(2,2),ED(3,2) *
C The third unit vector is then calculated by the *
C program as ED(.,3)=ED(.,1) X ED(.,2) *
C It may well be that a negative F has to be chosen *
C since the sign of F is effected by the direction of *
C the detector normal !!! Figure 1 in reference 2) may*
C be misleading. *
C AIR - Fraction of intensity loss per mm due to air *
C absorption. Usually AIR=0.001 . Each reflection *
C intensity is multiplied by *
C EXP(ABS(AIR*F*((1.0/COS(oblique angle))-1.0))), *
C the oblique angle is between diffracted beam and *
C detector normal and assumes values 0...<90 degrees. *
C 8 IGROUP- Space-group number of crystal. The numbers *
C corresponding to each possible space-group *
C are defined in the "INTERNATIONAL TABLES I". *
C At present, only the 65 enantiomorphic space groups *
C are implemented. A value of zero for IGROUP *
C indicates that space group and cell parameters *
C are unknown. *
C CELL - Unit cell parameters a,b,c (Angstroem) and *
C alpha,beta,gamma (degrees). *
C 9 MINUTE- Number of minutes the program should wait that an *
C expected data frame-file will be available. *
C FIXF - Refinement control variable *
C 0: refine detector to crystal distance, direct *
C beam direction and unit cell orientation. *
C The cell parameters are fixed. *
C 1: fix detector to crystal distance and refine *
C direct beam direction, unit cell orientation *
C and independent cell parameters. *
C DIVRAD- Half-angular size (degrees) of x-ray source *
C as seen from origin of coordinate system. *
C Approximately, DIVRAD=arctan(diameter of spot/2F) *
C DIVBET- Half-angular size (degrees) of mosaic spread *
C of crystal. Approximately, *
C DIVBET=0.5*(angular life-time of a spot passing *
C perpendicular through the Ewald sphere) *
C DELFRM- Maximum number of frames between Friedel-pairs *
C to be included to estimate anomalous dispersion *
C 10 REIDX - Integer array of 12 numbers providing a possibility *
C of reindexing the reflections in the final CORRECT *
C step. The meaning of the 12 numbers is defined as: *
C H' =REIDX(1)*H+REIDX( 2)*K+REIDX( 3)*L+REIDX( 4) *
C K' =REIDX(5)*H+REIDX( 6)*K+REIDX( 7)*L+REIDX( 8) *
C L' =REIDX(9)*H+REIDX(10)*K+REIDX(11)*L+REIDX(12) *
C where H',K',L' are the new indices. In case this *
C data card is omitted it is assumed that reindexing *
C is not required, that is H'=H, K'=K, L'=L . *
C *
C***********************************************************************
C *
C XDS.HKL (unformatted direct access) *
C ======= *
C *
C DESCRIPTION OF MAIN OUTPUT FILE *
C *
C The corrected reflection intensities are saved on this file in the *
C current directory. This file is an unformatted direct access file *
C of record length 68 bytes for each reflection. The file is sorted *
C with respect to the unique reflection indices. This means: *
C For each reflection with the original indices H,K,L all symmetry *
C equivalent indices are generated including Friedel related ones. *
C Among all these indices we choose the unique reflection indices *
C HA,KA,LA in the following order: HA is the largest H-index, among *
C those with the same HA-value select those with the largest K-index *
C which is KA, and finally the largest L-index which is called LA. *
C The unique indices HA,KA,LA thus found are packed into a 32-bit *
C word KEY=(LA+511)+(KA+511)*1024+(HA+511)*1048576 . *
C The reflections are then sorted in growing values of KEY. *
C *
C Record structure *
C *
C 16bit-WORD # CONTENTS *
C 1 HA (The last record is indicated by HA=10000) *
C 2 KA HA,KA,LA are the unique reflection indices. *
C 3 LA Any two reflections have the same unique *
C indices if and only if they are related by *
C symmetry. (HA,KA,LA are integer*2) *
C 4 H Original reflection indices H,K,L. *
C 5 K H,K,L are integer*2. *
C 6 L *
C 7 S Identifying number of symmetry operator used *
C to go from original to unique indices. *
C (integer*2). A negative sign indicates that *
C a mirror operation has been applied. This *
C information may be useful if a special *
C treatment for anomalous differences is *
C required which goes beyond the method of *
C the XDS-program. *
C 8 IPEAK Percentage of observed reflection intensity. *
C A value less than 100 indicates either a *
C reflection overlap or bad spots in the profile*
C 9 ICORR Percentage of correlation (integer*2) between *
C observed and expected reflection profile. *
C 10,11 FFADD LP-corrected intensity of this reflection *
C obtained by straight summation of counts *
C within spot region ( background subtracted). *
C The intensity is also corrected for radiation *
C damage and absorption. (real*4) *
C 12,13 SDADD Standard deviation of FFADD.(real*4) *
C 14,15 RLP Reciprocal LP-correction factor (real*4) *
C 16 ABSCAY Combined absorption and decay correction *
C factor*1000 (integer*2). *
C In case you want to remove this calculated *
C correction, divide intensities and standard *
C deviations by ABSCAY/1000.0 . *
C 17 IALFA IALFA and IBETA (both integer*2) are polar- *
C 18 IBETA coordinates of the spindle axis in units of a *
C hundreth of a degree. The lab coordinates of *
C the spindle axis are: *
C Ux=sin(BETA)*cos(ALPHA) *
C Uy=sin(BETA)*sin(ALPHA) *
C Uz=cos(BETA) *
C where ALPHA=IALFA/5729.578, *
C BETA =IBETA/5729.578. *
C 19 IFRM Frame number at diffraction of this reflection*
C (integer*2) *
C 20 PHI Calculated spindle position for this *
C reflection at diffraction in units of a *
C hundreth of a degree. (integer*2) *
C 21 IX, Calculated detector x- and y-coordinates for *
C 22 IY this reflection at diffraction in units of a *
C tenth of a pixel times 512.0/NX and 512.0/NY, *
C respectively. NX, NY are the numbers of pixels*
C along the detector X- and Y-axis. *
C IX,IY are integer*2. *
C 23-28 S0 Laboratory coordinates of direct beam wave- *
C vector ( rec. Angstroem). S0 points from the *
C x-ray source towards the crystal. *
C 29-34 S1 Laboratory coordinates of scattered beam wave-*
C vector. Length is 1.0/lambda (rec. Angstroem) *
C S0 and S1 are real*4 arrays of length 3. S1 *
C points from the crystal towards the detector. *
C At diffraction, laboratory coordinates of the *
C reflection H,K,L are: S1(.)-S0(.) *
C *
C***********************************************************************
C *
C UNIQUE.HKL (formatted sequential) *
C ========== *
C *
C DESCRIPTION OF SHORT OUTPUT FILE *
C *
C Symmetry related reflections are averaged and written with *
C FORMAT(3I5,4E12.4). Each record consists of *
C *
C HA,KA,LA,INTENSITY,STANDARD DEVIATION OF INTENSITY, *
C ANOMALOUS INTENSITY DIFFERENCE,STANDARD DEVIATION OF DIFFERENCE *
C *
C where HA,KA,LA are the unique reflection indices. The file is sorted *
C with respect to these unique reflection indices. The last record *
C is indicated by HA=10000. *
C *
C***********************************************************************
C *
C MISFITS (formatted sequential) *
C ======= *
C *
C Each line contains information about a rejected reflection which *
C has been recognized only by a comparison with symmetry equivalent *
C reflections. Format is: FORMAT(6I5,3E12.4,2I4,3I8) *
C *
C HA,KA,LA,H,K,L,FFADD,SDADD,RLP,FFRM,LFRM,XCAL,YCAL,PHICAL *
C *
C The definition of symbols is as described for file XDS.HKL *
C *
C***********************************************************************
C *
C SPOT.XDS (formatted sequential) *
C ======== *
C *
C The file contains a list of strong observed diffraction spots. *
C The spots are located by "COLSPOT" and written to file as *
C IX*10,IY*10,PHI*100,INTENSITY using FORMAT(4I7) *
C The spot positions IX,IY (pixels) are already corrected for *
C spatial distortion, PHI (degrees) is observed spindle position. *
C The file is sorted in decreasing spot intensities. *
C This file is read subsequently by "IDXREF" and indices are *
C attached to each spot. Indices 0 0 0 are used to indicate a spot *
C that could not be indexed. On return, the original file contents *
C is replaced by *
C IX*10,IY*10,PHI*100,INTENSITY,H,K,L using FORMAT(4I7,3I5) *
C for each spot, and H,K,L are the reflection indices. *
C In case of problems with the autoindexing routine it may be *
C useful to inspect this file and remove weak spots near the end. *
C *
C***********************************************************************
C *
C BKGPIX.XDS (unformatted direct access) *
C ========== *
C *
C Record structure *
C The file contains NY records each of length 2*NX bytes, where *
C NX,NY are the number of pixels along the detector X- and Y-axis. *
C *
C Contents *
C The file contains an image of the array : INTEGER*2 IBKG(NX*NY) *
C such that IBKG(IX+NX*(IY-1)) is the sum of background values at *
C pixel position IX,IY (0