cp2k/tests/SE/regtest/O2-ROKS.inp

&GLOBAL
  PRINT_LEVEL LOW
  PROJECT O2-ROKS
  RUN_TYPE GEO_OPT
&END GLOBAL

&MOTION
  &GEO_OPT
    MAX_DR 0.001
    MAX_FORCE 0.00015
    MAX_ITER 30
    OPTIMIZER BFGS
    RMS_DR 0.0005
    RMS_FORCE 0.0001
  &END GEO_OPT
  &PRINT
    &STRUCTURE_DATA
      DISTANCE 1 2
    &END STRUCTURE_DATA
  &END PRINT
&END MOTION

&FORCE_EVAL
  METHOD Quickstep
  &DFT
    MULTIP 3
    ROKS
    &QS
      METHOD MNDO
      &SE
        ANALYTICAL_GRADIENTS F
      &END SE
    &END QS
    &SCF
      EPS_SCF 1.0E-6
      SCF_GUESS atomic
    &END SCF
  &END DFT
  &SUBSYS
    &CELL
      ABC 10.0 10.0 10.0
      PERIODIC NONE
    &END CELL
    &COORD
      O     0.000000     0.000000    -0.600000
      O     0.000000     0.000000     0.600000
    &END COORD
    &PRINT
      &KINDS
        BASIS_SET
        POTENTIAL
      &END KINDS
      &STRUCTURE_DATA
        DISTANCE 1 2
      &END STRUCTURE_DATA
    &END PRINT
  &END SUBSYS
&END FORCE_EVAL
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&GLOBAL`
			`PRINT_LEVEL LOW`
a more appropriate choice for PROJECT_NAME : use different ones unless a RESTART is in place. svn-origin-rev: 10814 2010-12-28 13:33:01 +00:00			`PROJECT O2-ROKS`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`RUN_TYPE GEO_OPT`
			`&END GLOBAL`
Format all input files 2023-12-18 20:55:03 +01:00
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&MOTION`
rename GEOOPT section in GEO_OPT (consistent with RUN_TYPE GEO_OPT) svn-origin-rev: 6375 2007-09-26 15:23:48 +00:00			`&GEO_OPT`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`MAX_DR 0.001`
			`MAX_FORCE 0.00015`
			`MAX_ITER 30`
			`OPTIMIZER BFGS`
			`RMS_DR 0.0005`
			`RMS_FORCE 0.0001`
rename GEOOPT section in GEO_OPT (consistent with RUN_TYPE GEO_OPT) svn-origin-rev: 6375 2007-09-26 15:23:48 +00:00			`&END GEO_OPT`
Mainly getting ready for FE-PQA: (1) A more logical setup for multiple_force_eval.. In order to save memory and also to avoid a duplication of information it is possible to specify just one single subsys when using multiple_force_eval (or different).. This has been possible cleaning all referential calls to subsys.. The code now is definitely more structured and cleaned. At the same time several cleanings have been performed while I was working on the single files. In particular structure_data has been moved to the print section of the motion section. Same for printing subcell.. moved to the respective section in MM and DFT. This reflects the idea of making logical partitions in the input. Heavy lifting around (not just cosmetics).. regtest ok! svn-origin-rev: 5942 2007-04-15 22:07:57 +00:00			`&PRINT`
			`&STRUCTURE_DATA`
			`DISTANCE 1 2`
			`&END STRUCTURE_DATA`
			`&END PRINT`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&END MOTION`
Format all input files 2023-12-18 20:55:03 +01:00
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&FORCE_EVAL`
RESPA: change in the input structure &GLOBAL PROGRAM -> &FORCE_EVAL METHOD (see regtests). Two fully independent FORCE_EVAL sections can be around in the input. Thus, do not directly access them from the globenv%input_file, but use the force_env(section) directly. force_env section now contains a flag indicating what method it describes. All now incorrect uses of globenv%input_file fixed. svn-origin-rev: 5283 2006-08-08 10:50:22 +00:00			`METHOD Quickstep`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&DFT`
			`MULTIP 3`
Format all input files 2023-12-18 20:55:03 +01:00			`ROKS`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&QS`
			`METHOD MNDO`
Tsunami effects (1): Analytical gradients for semiempirical methods. Nuclear gradients are since now on computed analytically. Also the value of the integral for the limiting case of two atoms aligned in the z direction is computed EXACTLY without performing the limit. The old situation can be restored disablign the keyword ANALYTICAL_GRADIENTS in the SE section. New regtests. Resetting few qm/mm SE because of the use of analytical expressions. waiting for the next wave... svn-origin-rev: 5930 2007-04-11 08:49:00 +00:00			`&SE`
			`ANALYTICAL_GRADIENTS F`
			`&END SE`
ROKS for semiempirical methods (SE), i.e. for orthogonal basis sets in general svn-origin-rev: 5131 2006-05-09 10:20:28 +00:00			`&END QS`
			`&SCF`
			`EPS_SCF 1.0E-6`
			`SCF_GUESS atomic`
			`&END SCF`
			`&END DFT`
			`&SUBSYS`
			`&CELL`
			`ABC 10.0 10.0 10.0`
			`PERIODIC NONE`
			`&END CELL`
			`&COORD`
			`O 0.000000 0.000000 -0.600000`
			`O 0.000000 0.000000 0.600000`
			`&END COORD`
			`&PRINT`
			`&KINDS`
			`BASIS_SET`
			`POTENTIAL`
			`&END KINDS`
			`&STRUCTURE_DATA`
			`DISTANCE 1 2`
			`&END STRUCTURE_DATA`
			`&END PRINT`
			`&END SUBSYS`
			`&END FORCE_EVAL`