{+ file: make_cv_twin.inp +}
{+ directory: xtal_twin +}
{+ description: Setup test array for cross-validation (free R) using a
                random selection data with hemihedral twinning  +}
{+ comment:
            This script MUST be used for refinement of twinned structures.
            Friedel mates are conserved for anomalous data.
            The test set array can also be extended if it already 
            exists in the data set.  In this case, a resolution range
            or selection should be specified for the generation of 
            the test set. +}
{+ authors: Axel T. Brunger and Paul D. Adams +}
{+ copyright: Yale University +}

{+ reference: A.T. Brunger, The Free R Value: a Novel Statistical 
              Quantity for Assessing the Accuracy of Crystal Structures, 
              Nature 355, 472-474 (1992) +}
 
{- Guidelines for using this file:
   - all strings must be quoted by double-quotes
   - logical variables (true/false) are not quoted
   - do not remove any evaluate statements from the file -}

{- begin block parameter definition -} define(

{====================== crystallographic data ========================}

{* space group *}
{* use International Table conventions with subscripts substituted
   by parenthesis *}
{===>} sg="R3";

{* unit cell parameters in Angstroms and degrees *}
{+ table: rows=1 "cell" cols=6 "a" "b" "c" "alpha" "beta" "gamma" +}
{===>} a=104.400;
{===>} b=104.400;
{===>} c=124.250;
{===>} alpha=90;
{===>} beta=90;
{===>} gamma=120;

{* reciprocal space array containing test set for cross-validation *}
{* this array will be created if it does not already exist and 
   defined for the selected reflections.  If it exists the 
   test set will be (re)defined for the selected reflections. *}
{===>} test_set="test";

{* number for selection of test reflections: required *}
{* ie. reflections with the test set array equal to this number will be
       used for cross-validation, all other reflections form the working set *}
{===>} test_flag=1;

{* reflection file - to which test set array will be added *}
{===>} reflection_infile="porin.hkl";

{======================= twinning parameters =========================}

{* twinning operation *}
{===>} twin_oper="h,-h-k,-l";

{======================= test set generation ==========================}

{* To generate a new test set, use the whole resolution range and
   select all reflections.  To extend a test array (e.g. to a higher 
   resolution data set), use the xtal_util/merge.inp task file 
   to merge the test set and the new diffraction data, and then use 
   this task file to extend the test set. *}

{* percentage of reflections to be used for TEST set - ie. 10% *}
{===>} percentage=10.0;

{* Resolution limits for generation of test set *}
{* Normally, the full resolution range (or larger) of observed data
   should be used. Use a restricted set when extending an existing test
   array (e.g, to high resolution). *}
{+ table: rows=1 "resolution" cols=2 "lowest" "highest" +}
{===>} low_res=500.0;
{===>} high_res=1.0;

{* Additional selection for generation of test set *}
{* Normally, select all reflections.  Specify this additional selection 
   if the test set needs to be expanded for specific subset(s) of 
   reflections. *}
{===>} ref_sel=(all);

{=========================== output files ============================}

{* output reflection file *}
{===>} reflection_outfile="make_cv_twin.cv";

{===========================================================================}
{        things below this line do not normally need to be changed          }
{===========================================================================}

 ) {- end block parameter definition -}

 checkversion 1.2

 evaluate ($log_level=quiet)

 xray

   @CNS_XTALLIB:spacegroup.lib (sg=&sg;)

   a=&a b=&b c=&c  alpha=&alpha beta=&beta gamma=&gamma 

   reflection @&reflection_infile end 

   if (&test_flag=0) then
      evaluate ($work_flag=-1) 
   else
      evaluate ($work_flag=0)
   end if
     
   set echo=off end 
   if ( &BLANK%test_set = true ) then
     display 
     display  ***********************************
     display  Error: test set array not specified
     display  ***********************************
     display
     abort
   else
      query name=&STRIP%test_set domain=reciprocal end
      if ( $object_exist = false ) then
         declare name=&STRIP%test_set domain=reciprocal type=integer end
         do ( &STRIP%test_set = $work_flag ) ( all )
      else
         {- test array can be of type INTEGER or REAL -}
         if ( $object_type = "COMPLEX" ) then
            display 
            display ********************************************************
            display Error : test set array &test_set has the wrong data type
            display ********************************************************
            display
            abort
        end if
      end if
   end if
   set echo=on end

   declare name=twin_sel domain=reciprocal type=integer end

   do ( twin_sel=0 ) (all)
   do ( twin_sel=1 ) ( &low_res >= d >= &high_res and &ref_sel )
   do ( twin_sel=min(twin_sel,abs(remap[&STRIP%twin_oper](twin_sel))) ) 
                         ( &low_res >= d >= &high_res and &ref_sel )

   do ( &STRIP%test_set = $work_flag )
      ( &low_res >= d >= &high_res and &ref_sel ) 
   
   anomalous ?
   if ($result=FALSE) then
      do (&STRIP%test_set=&test_flag*int(random()+&percentage/200.)) 
             ( &low_res >= d >= &high_res and &ref_sel and twin_sel=1 ) 
   else
      {- make sure both Friedel mates are included for each test reflection -}
      do (&STRIP%test_set=&test_flag*int(random()+&percentage/200.))  
             ( centric  and &low_res >= d >= &high_res and &ref_sel and twin_sel=1 )
      do (&STRIP%test_set=&test_flag*int(random()+&percentage/400.))  
             ( acentric and &low_res >= d >= &high_res and &ref_sel and twin_sel=1 )
      do (&STRIP%test_set=max(&STRIP%test_set,friedel(&STRIP%test_set)))
             ( acentric and &low_res >= d >= &high_res and &ref_sel and twin_sel=1 ) 
   end if   

   {- make sure all twin related reflections are in the same set -}
   do (&STRIP%test_set=max(&STRIP%test_set,
                           abs(remap[&STRIP%twin_oper](&STRIP%test_set)))) 
          ( &low_res >= d >= &high_res and &ref_sel and twin_sel=1 )

   undeclare name=twin_sel domain=reciprocal end

   write reflection 
      output=&reflection_outfile
   end  

 end

 stop