Controlling printed Output

The amount of printed output is regulated with the keys Print, NoPrint, EPrint and Debug. (No)Print and Debug are simple keys, EPrint is a block type key.

Many print options pertain to debugging situations and are included here only for completeness. This section is intended to give a survey of all possibilities. Some items may be mentioned again in other sections where the subject of a particular print switch is discussed.


The key DEBUG is used to generate extensive output that is usually only relevant for debugging purposes. It operates exactly like the PRINT key but there is no converse: nodebug is not recognized; it would be irrelevant anyway because by default all debug print switches are off.

A list of the possible items for the DEBUG key is given below.

All items of the print list can also be used with the debug key. If they are not mentioned in table III, the meaning is the same as for the print key, but the corresponding output may be generated more often, for instance at every SCF cycle rather than at the last one only.

Table III. Arguments for the print key DEBUG. All debug switches are by default off.
Item Explanation
Basis Construction of the orthonormal LOW basis from elementary (BAS) and fragment (FO) basis.
Core Core Orthogonalization procedure
Ekin Kinetic energy matrices. (compare the print switches EKIN)
Fit Construction of the symmetry adapted fit functions
Fitint Construction of integrals used in the Fit procedure.
Freq Force matrices processed in the computation of frequencies:
  Cartesian and internal representation, before and after symmetrization,
  etc. (as far as applicable).
GeoStep Geometry optimization procedure. All relevant items.
Hess Complete eigensystem of the Hessian during geometry optimizations.
NumInt Numerical integration. Very extensive output
  (including the coordinates and weights of all generated points).
Pmat P-matrix (density matrix) during SCF
  and in the ETS analysis program in the BAS representation.
Rhofih Computation of fit coefficients during the SCF.
SCF Extensive output during the SCF procedure about many different items.
  See also EPRINT, subkey SCF.
SDIIS All data concerning the DIIS as used during the SCF.
  See ERPRINT, subkey SDIIS.
TransitionField The Transition State procedure to compute and analyze certain terms in the bonding energy.
  The distinct components, the involved transition field Fock matrices, etc.


The key EPRINT is an extended version of the (no)print key, employed for print switches that require more specification than just off or on.

Contrary to what is the case for the keys print and noprint, the key EPRINT must occur only once in the input file; any subsequent occurrences are incorrect and ignored or lead to abort.

A subkey-type structure: it consists of a keyword followed by data, so that it functions as a simple (sub)key, or it is a keyword followed by a data block which must then end with the word subend.

The subkeys used in the EPRINT data block are called Eprint keys. A complete list of them is given below. All available EPRINT keys are discussed in the schemes below. The enclosing records EPRINT and end are omitted in these schemes.

Table IV. List of EPRINT subkeys.
EPRINT subkeys Subject
AtomPop Mulliken population analysis on a per-atom basis
BASPop Mulliken population analysis on a per-bas-function basis
Eigval One-electron orbital energies
Fit Fit functions and fit coefficients
Frag Building of the molecule from fragments.
FragPop Mulliken population analysis on a per fragment basis
Freq Intermediate results in the computation of frequencies (see debug: freq).
GeoStep Geometry updates (Optimization, Transition State, ...)
NumInt Numerical Integration
OrbPop (Mulliken type) population analysis for individual MOs
OrbPopEr Energy Range (ER) in hartree units for the OrbPop subkey
Repeat repetition of output in Geometry iterations (SCF, optimization, ...)
SCF Self Consistent Field procedure
SFO Information related to the Symmetrized Fragment Orbitals and the analysis
TF Transition Field method.

Eprint subkeys vs. Print switches

Several EPRINT subkeys are merely shortcuts for normal (no)print switches. All such simple subkeys are used in the following way:

  ESUBKEY argumentlist
One of the following EPRINT subkeys: Fit, Frag, GeoStep, NumInt, Repeat, SCF, sdiis, SFO, TF, Time.
A sequence of names, separated by delimiters. Each of these names will be concatenated with the esubkey and the combination will be stored as a normal print switch. Example: Frag rot, SFO will be concatenated to fragrot and fragsfo and both will be stored as print switches. All such combinations can also be specified directly with the key PRINT. The example is therefore exactly equivalent with the input specification: print FragRot, Fragsfo****

If any of the names starts with the two characters no, the remainder of the name will be concatenated with the EPRINT, but now the result will be stored and treated as a noprint switch. Items that are on by default can in this way be turned off. Example:

  FRAG noRot Eig

This turns Rot off and Eig on for the EPRINT subkey Frag. Equivalent would be:

Print FragEig

Follows a description of all simple EPrint subkeys:


The subkey fit controls output of how the elementary fit functions are combined into the symmetric (A1) fit functions. It controls also printing of the initial (start-up) and the final (SCF) fit coefficients.

  FIT list
A list of items, separated by blanks or commas. The following items are recognized: Charge, Coef, Comb.
The amount of electronic charge contained in the fit (start-up), total and per fragment.
The fit coefficients that give the expansion of the charge density in the elementary fit functions.

The construction of the totally symmetric (A1) fit function combinations from the elementary fit functions.

By default all options are off.


The subkey frag controls output of how the molecule is built up from its fragments.

  FRAG list
A list of items, separated by blanks or commas. The following items are recognized: Eig, Fit, Rot, SFO.
The expansion coefficients in elementary functions (bas) of the fragment Molecular Orbitals as they are on the fragment file.
The rotation (and translation) required to map the master fragment (i.e. the geometrical data on the fragment file) onto the actual fragment which is part of the current molecule. N.B.: if eig and rot are both on, the rotated fragment orbitals are printed also.
The fit coefficients that describe the fitted charge density of the fragments after the rotation from the master fragment on file to the actual fragment. These are the molecular fit coefficients that are used (by default) to construct the total molecular start-up (fitted) charge density and hence the initial Coulomb and XC potential derived from it.
The Symmetry-adapted combinations of Fragment Orbitals that are used in the current calculation. This feature ensures that the definition of the SFOs is printed. This will happen anyway whenever the EPRINT subkey SFO itself is activated. By default all options are off. Remark: SFO analysis in a Spin-Orbit relativistic calculation is implemented only in the case there is one scalar relativistic fragment., which is the whole molecule.


Controls printing of Force matrices and a few more data that are intermediate results in the computation of frequencies after all coordinate displacements have been carried out.

  FREQ list
contains any of the items SymCoord, DMuRot, Hess.
print the Symmetry Coordinates, both as they are generated, and some related info in their processing later on. The Symmetry Coordinates are symmetry-adapted combinations of Cartesian displacements, with the pure translations and rotations projected out.
info about generating Dipole-Derivative information in transformation between Cartesian and internal coordinate representation as regards the rotational aspects.

processing of the completed matrix of force constants, symmetrization, transformation to other coordinates.

By default all options are off.

Controls output concerning the geometry update method, parameters, energy gradients, etc. It plays no role in a SinglePoint calculation.



A list of items, separated by blanks or commas. The following items are recognized: Energy, GradientTerms, Gradients, Upd.
summary of the (bonding) energy and its components as computed in the geometry update procedure.
Energy gradients on the free variables. These may be all or some of the Cartesian or the Z-matrix coordinates, depending on the case.
The decomposition of the gradients in computed terms, as described in the thesis of L.Versluis [7].

parameters used and adapted in the geometry update procedure.

By default Gradients, Upd are on, the other items off.


Output related to the numerical integration procedure: parameters, generated points, tests on the accuracy of the generated scheme, etc.

  NUMINT list
A list of items, separated by blanks or commas. The following items are recognized: All, Geo, Ovl, Par, Pnt, Res, Sym, Test.
includes all other options and prints in addition the coordinates and weights of all generated points.* This can be a lot of output!*
geometric data such as boundary planes around the molecule, as they are computed and used in the program section where the point grid is generated.
numerically integrated are the auto-overlaps of symmetry-adapted combinations of elementary basis functions SBAS. The deviations from the analytically computed values is printed. The test option, see below, yields a summary of these data: the maximum error and the root-mean-square error.
employed precision parameters, atomic spheres radii etc.
the generated numbers of points in each of the subregions processed in the point-generating procedure.
results as regards the total number of points, the sum-of-weights and the partitioning of the points in blocks (for segmented vectorization).
the symmetry operators that are computed directly from the coordinates (irrespective of the input Schönfliess symbol) and that are used to construct thec numerical integration grid in a symmetric fashion.

a few external tests are performed after the grid has been generated, such as the numerical integration of the sum-of-fragment densities. See also the norms option.

By default Res and Test are on, the other options off.


Specifies that (Mulliken type) population analysis should be printed for individual MOs, both on a per-SFO basis and on a per-bas function basis. The format of the subkey is as follows:

  ORBPOP TOL=X Nocc Nunocc

X is the threshold for the SFO coefficient value to include in the listing for the per-SFO analysis. Nocc is the number of the highest occupied and Nunocc is the number of the lowest unoccupied orbitals to analyze.


Specifies the energy range for the MOs to which the OrbPop key applies. The default range is from -0.7 below the HOMO to 0.2 Hartree above the LUMO. Usage:

  OrbPopER minEn maxEn

where minEn and maxEn are both in Hartree, and have the defaults just specified. In order to get information on many more orbitals, simply specify a large negative value for minen and a large positive value to maxen.


Control the repetition of output in Geometry iterations: optimization, computation of frequencies, transition state search.

  Repeat list
contains one or more of the following items: NumInt, SCF.
Output from the numerical integration procedure, like parameters, numbers of points generated, test data is controlled by the numint subkey (see below). The repeat subkey controls whether the output is repeated for all geometries (if the flag is on) or only for the first (if the flag is off). Some concise info is produced (repeatedly) anyway if the print switch computation is on.

Controls similarly the SCF output, like population analysis and orbital eigenvalues. If the flag is on, these items are printed at the last SCF cycle in every geometry, otherwise only at the last (in case of an optimization, not in case of a Frequencies calculation).

By default both options are off.


Output during the SCF procedure.

  SCF list
is a list of items, separated by blanks or commas. The following items are recognized: Eigval, Eigvec, Err, Fmat, Keeporb, MOPop, Occ, Pmat, Pop, Start.
Eigenvalues of the one-electron orbitals at the last SCF cycle. In a run with multiple SCF runs (Geometry Optimization,..) this printing occurs only for the last SCF procedure. See also the eigval subkey of EPRINT. (Use debug or the repeat subkey of EPRINT to get output on all cycles).
MO eigenvector coefficients in the BAS representation. Only printed on the last SCF cycle.
SCF error data which are checked for convergence. By default this takes effect after cycle 25 of the SCF. If the key is set it takes effect at the first cycle. Optionally one may type ErrN,where n is an integer (written directly after Err without a blank in between), in which case the key takes effect at cycle n.
Fock matrix in the low representation.
If the KeepOrbitals option is activated (see the key SCF), output is generated whenever this option actually results in a change of occupation numbers as regards the energy ordering.
concise output of SCF occupation numbers on last SCF cycle if no eigenvalues are printed (see: Eigval).
Mulliken populations in terms of the elementary basis functions (bas), per MO, for input-specified MOs (see the EPRINT subkey orbpop)
Density matrix
General control of bas Mulliken populations. This supervises all printing (whether populations are printed or not) according to the EPRINT subkeys atompop, fragpop, orbpop (the latter only as regards the bas population analysis at the end of the SCF procedure).
Data pertaining to the first SCF cycle (of the first SCF procedure, in case of an optimization; use repeat to get this for all SCFs).

By default Eigval, Keeporb, Occ, and Pop are on, the others off.


Information pertaining to the use of Symmetrized Fragment Orbitals (for analysis purposes).

  SFO list
A list of items, separated by blanks or commas. The following items are recognized: eig, eigcf, orbpop, grosspop, fragpop, ovl.
The MO coefficients in terms of the SFOs.
idem, but now also containing the coefficients pertaining to the CoreFunctions.
population analysis of individual orbitals. The orbitals analyzed are set with the EPRINT subkey orbpop.
Gross populations of the SFOs, split out in symmetry representations. GrossPop is automatically turned on when OrbPop is activated.
Population analysis on a per-FragmentType basis. This analysis does in fact not depend on the SFOs (ie, the result does not depend on how the SFOs are defined), but the computation of these populations takes place in the SFO-analysis module, which is why it is controlled by the SFO print option. FragPop output is given per orbital when OrbPop is activated, per symmetry representation when GrossPop is activated, and as a sum-over-all-orbitals-in-all-irreps otherwise (if FragPop is active).

Overlap matrix of the SFO basis, separately for each symmetry representation.

By default orbpop is on, the other options off.

In a Spin-Orbit calculation the SFO analysis is not yet implemented completely.

Remark: the options eig and eigcf replace the previous (now disabled) simple print options eigsfo and eigsfo.

Note that the simple print key SFO controls whether or not the EPRINT subkey sfo is effective at all.


Part of the bonding energy is computed and analyzed by the so-called Transition State procedure [3, 110]. This has nothing to do with physical transition states, but is related to the Fock operator defined by an average charge density, where the average is taken of the initial (sum-of-orthogonalized-fragments) and the final (SCF) charge density. There is also an analogous term where the average is taken of the sum-of-fragments and the sum-of-orthogonalized-fragments. Various terms, Fock operators and Density Matrices used in this approach may be printed. To avoid confusion with real Transition States (saddle points in the molecular Energy surface) the phrase TransitionField is used here.

  TF list
A list of items, separated by blanks or commas. The following items are recognized: Energy, Fmat, DiagFmat, FragPmat, DiagFragPmat, F*dPmat, DiagF*dPmat, OrbE.
Energy terms computed from the TransitionField.
TransitionField Fock matrices.
Idem, but only the diagonal elements.
The molecular P-matrix constructed from the sum-of-fragments.
idem, but only the diagonal elements.
The TransitionField energy term can be expressed as a Fock operator times the difference between two P-matrices (initial and final density).
only diagonal elements
Orbital energies in the TransitionField.

By default all options are off.

Other Eprint subkeys

We discuss now the remaining EPRINT sub keys that are not simple shortcuts for print switches.

  Eigval noccup {nvirtual}

This specifies the number of one-electron orbitals for which in the SCF procedure energies and occupation numbers are printed whenever such data is output: the highest noccup occupied orbitals and the lowest nvirtual empty orbitals. Default values are noccup=10, nvirtual=10. If only one integer is specified it is taken as the noccup value and nvirtual is assumed to retain its standard value (10). Printing can be turned off completely with the EPRINT sub key SCF, see above.

Mulliken Population Analysis

All population subkeys of EPRINT refer to Mulliken type populations.

  ATOMPOP level

Populations accumulated per atom.

level must be none, gross or matrix. none completely suppresses printing of the populations; gross yields the gross populations; matrix produces the complete matrix of net and overlap populations. Default value: matrix.

  BASPop level

Populations are printed per elementary (bas) basis function. The level options are none, short, gross, matrix. none, gross and matrix are as for atompop.

short yields a summary of BAS gross populations accumulated per angular momentum (l) value and per atom.

Default value: gross.

  FragPop level

Completely similar to the atompop case, but now the populations per fragment. Of course in the case of single-atom fragments this is the same as atompop and only one of them is printed. Default: matrix.

For all three population keys atompop, fragpop and baspop, specification of a higher level implies that the lower-level data, which are in general summaries of the more detailed higher level options, are also printed.

Printing of any populations at the end of the SCF procedure is controlled with the EPRINT sub key SCF (pop).

Population Analysis per MO

A very detailed population analysis tool is available: the populations per orbital (MO). The printed values are independent of the occupation numbers of the MOs, so they are not populations in a strict sense. The actual populations are obtained by multiplying the results with the orbital occupations.

The analysis is given in terms of the SFOs and provides a very useful characterization of the MOs at the end of the calculation, after any geometry optimization has finished. This feature is now also available in a Spin-Orbit coupled relativistic calculation, in the case there is one scalar relativistic fragment, which is the whole molecule.

The same analysis is optionally (see EPRINT subkey SCF, option mopop also provided in terms of the elementary basis functions (bas).

  OrbPop {noccup {nvirtual}} {tol=tol}
    subspecies orbitals
    subspecies orbitals
Determines how many of the highest occupied orbitals are analyzed in each irrep. Default noccup=10.
Determines in similar fashion how many of the lowest virtual orbitals are analyzed in each irrep. Default nvirtual=4.
Tolerance parameter. Output of SFO contributions smaller than this tolerance may be suppressed. Default: 1e-2.
One of the subspecies of the molecular symmetry group. Can not be used (yet) in a Spin-Orbit coupled calculation.
A list of integers denoting the valence orbitals (in energy ordering) in this subspecies that you want to analyze. This overrules the noccup,nvirtual specification for that symmetry representation. In an unrestricted calculation two sequences of integers must be supplied, separated by a double slash (//).

Any subset of the subspecies can be specified; it is not necessary to use all of them. No subspecies must occur more than once in the data block. This can not be used in a Spin-Orbit coupled equation (yet).

A total SFO gross populations analysis (from a summation over the occupied MOs) and an SFO population analysis per fragment type are preformed unless all MO SFO-populations are suppressed.

Reduction of output

One of the strong points of ADF is the analysis in terms of fragments and fragment orbitals (SFOs) that the program provides. This aspect causes a lot of output to be produced, in particular as regards information that pertains to the SFOs.

Furthermore, during the SCF and, if applicable, geometry optimizations, quite a bit of output is produced that has relevance merely to check progress of the computation and to understand the causes for failure when such might happen.

If you dislike the standard amount of output you may benefit from the following suggestions:

If you are not interested in info about progress of the computation:

NOPRINT Computation

If you’d like to suppress only the SCF-related part of the computational report and make the GeometryUpdates related part more concise:


(Keep computation on, so you get at least some info about the GeometryUpdates)

If you don’t want to see any SFO stuff:


To keep the SFO definitions (in an early part of output) but suppress the SFO-mo coefficients and the SFO overlap matrix:

 SFO noeig, noovl

Note: the SFO-overlap matrix is relevant only when you have the SFO-MO coefficients: the overlap info is needed then to interpret the bonding/anti-bonding nature of the various SFO components in an MO.

If you are not interested in the SFO populations:

 SFO noorbpop