The structure of a typical dataset would be of the form
| RUN PARAMETERS | (Optional) e.g. MEMORY, TIME |
| TITLE | One-line description of your choosing. |
| GEOMETRY PARAMETERS | Relevant parameters related to the |
| geometry such as the directives | |
| ANGSTROM, SYMMETRY, | |
| CONSTANTS, VARIABLES, | |
| BASIS, DIHEDRAL . | |
| GEOMETRY SPECIFICATION | |
| WAVEFUNCTION SPECIFICATION | |
| ALL OTHER DIRECTIVES | |
| START | indicates that the calculation is to start |
| FINISH | Signals the end of the dataset. |
TITLE NH2 631G* uhf SYMMETRY CNV 2 END BASIS 631G* VARIABLES NH 1.01 A ZNH 50.0 D END ATOMS Nitrogen 7 0.0 0.0 0.0 Hydrogen 1 POL NH ZNH END UHF MULTIPLICITY 2 ENERGY START FINISHThis contains various directives explained later, but the structure corresponds to the general form given above - there are parameters specifying the geometry, and the basis set. The geometry in this example is in polar coordinates which is only useful for small molecules, and a more general internal coordinate input is described later. It is also possible to use simple cartesian geometries. The dataset also contains the type of wavefunction (UHF) and the type of calculation (ENERGY). The calculation commences when the START directive is encountered, and finishes when FINISH is found. If you insert further commands between the START and FINISH then further calculations continuing from the end of the energy calculation will occur. Note the use of END to terminate multiline directives. Two further points about datasets in general - all input is in free format and all commands must be in upper case. We shall now work through the possible commands in roughly the order they occur in the example dataset.