ADFprepare ($ADFBIN/adfprep) generates a job script from a .adf file (the template). The .adf file can be produced by ADFinput, or you can use one of the default templates included. These default templates are identical to those present in ADFinput.
Two examples are included to give you an idea what you can do with adfprep.
In $ADFHOME/examples/adf/BakersetSP you will find how to use adfprep to run a particular job (a single point calculation in this case) for all molecules in the Baker set. The molecules are simply xyz files and contain no ADF specific information. adfreport is used to collect the resulting bonding energies
In $ADFHOME/examples/adf/ConvergenceTestCH4 you will find how to use adfprep to test convergence of the bonding energy with respect to basis set and integration accuracy. adfreport is used to collect the resulting bonding energies
The most convenient way to see the optoins of adfprep is to run the adfprep command without arguments. You will get output very much alike the following description, but probably more up-to-date.
Usage: adfprep -t template.adf [-m molecule.(adf|xyz|mol|t21)] [-z charge] [-s spin] [-i integration]
[-b basis] [-c core] [-r relativity] [-x xcpotential] [-e xcenergy]
[-j jobname] [-a adffile]
Start with a template, adjust it for this particular job, and write the resulting job to standard output.
Values specified should match exactly the values as you would specify using ADFinput, also for menu choices.
TEMPLATE
-t: the .adf file (saved by ADFinput) to be used as template, defining the whole job
All other options override values from this job
Instead of a .adf file, you may also specify the name of one of the standard templates as defined in
ADFinput: "Single Point", Frequencies, "Geometry Optimization", etc
Some shortcuts: SP, GO, GOS, FREQ
CHANGES TO TEMPLATE
-m: the molecule to use, element types and coordinates
This can be taken from anything that ADFinput can import, for example .adf, .mol, xyz or .t21 files
-z: charge (real number)
-s: spin (integer), if not zero this implies an unrestricted calculation
-i: integration (integer)
-b: basis type (SZ, DZ, DZP, TZ, TCP, TZ2P, QZ4P)
-c: core type (None, Small, Medium, Large)
-r: relativistic level (None, Scalar, Spin-Orbit), using ZORA
-x: XC potential during SCF, one from the options available in ADFinput:
LDA,
GGA:BP, GGA:BLYP, GGA:PW91, GGA:mPW, GGA:PBE, GGA:RPBE, GGA:revPBE, GGA:mPBE, GGA:OLYP, GGA:OPBE,
Model:SAOP, Model:LB94,
Hartree-Fock,
Hybrid:B3LYP, Hybrid:B3LYP*, Hybrid:B1LYP, Hybrid:KMLYP, Hybrid:O3LYP, Hybrid:X3LYP, Hybrid:BHandH,
Hybrid:BHandHLYP, Hybrid:B1PW91, Hybrid:MPW1PW, Hybrid:MPW1K, Hybrid:PBE0, Hybrid:OPBE0
-e: XC energy after SCF (Default, LDA+GGA_METAGGA, LDA+GGA+METAGGA+HYBRIDS)
-k: replace any key, the single argument will be broken into:
the key, the replacement value, and END for a block key
all separated by spaces. To insert a return, add a |
When the key is not found, it is added just before the ATOMS key
The -k key may be repeated, and is applied at the end, replacing even earlier changes
OUTPUT
-j: produce a fully runnable job (as the .job files from ADFjobs), using the specified jobname
the job script produces files like jobname.out, jobname.t21 etc. Several job scripts can simply
be concatenated, the results will be stored in different files using th jobname parameter
the default is a simple run script (the .run file from ADFinput, files are left as they are)
-a: save a .adf file that matches the run script, except for the -k arguments (they are listed in the user
input field) adffile is the name of the adffile, including the .adf extension (required)
