Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision		 Previous revision
recent_changes [2025/02/06 08:33] – fix broken link mayrecent_changes [2025/10/25 07:22] (current) – [New Option CCF12 in PNO-LCCSD(T)-F12] werner
Line 5: Line 5:
 We recommend always to use the most recent version, since developments are ongoing and problems reported by users are always fixed as quickly as possible. In particular, before reporting bugs, please check if these still occur in the latest version. We recommend always to use the most recent version, since developments are ongoing and problems reported by users are always fixed as quickly as possible. In particular, before reporting bugs, please check if these still occur in the latest version.
  
-=====  Upcoming: GUI gmolpro 2.x.x =====+ 
 + 
 +===== New features of MOLPRO2025.4 ===== 
 + 
 +Bugfixes 
 + 
 +==== Simplfied Input ==== 
 + 
 +A simplified one-line input for standard calculations has been added, see [[Quickstart|QuickStart]]. 
 + 
 +==== New Option CCF12 in PNO-LCCSD(T)-F12 ==== 
 + 
 +A new option has been added to the PNO program, see [[Local correlation methods with pair natural orbitals (PNOs)#Options for F12 calculations|Options for F12 calculations in PNO-LCCSD(T)-F12]]. With CCF12=0  F12 is not included in the PNO-LCCSD step and the MP2 F12 correction is added to the PNO-LCCSD and PNO-LCCSD(T) energies. This is similar to a ''focal point'' approximation, where the high-level calculation is PNO-LMP2-F12 and the low-level one PNO-LCCSD(T). 
 +===== New features of MOLPRO2025.3 ===== 
 + 
 +==== CABS single correction of the HF energy ==== 
 + 
 +The program for the CABS singles correction has been improved and now runs fine in parallel also on multiple nodes. Unless already available, the cabs correction will from now on be automatically computed in the PNO-F12 programs. It is stored in the orbital dump record and reused if available for the given orbital basis.  
 + 
 +In large cases the non-parallel diagonalizations needed for basis orthonormalisation and block-diagonalization of the Fock matrix may take more time than the Fock matrix evaluation. To reduce the diagonalization time, the PNO program uses by default ORTHO_CABS=0. This saves one large diagonalization, but has only a negligible effect on relative energies. This default can be overwritten in the PNO program by setting option ortho_cabs=1 (which has been and still is the default if the cabs correction is computed with df-cabs or in non-PNO F12 calculations). 
 + 
 + 
 +Please see section  section [[Local correlation methods with pair natural orbitals (PNOs)|Local correlation methods with pair natural orbitals (PNOs)]] for further details, e.g. how to disable this mechanism or how to force a recalculation of the CABS correction. 
 + 
 +==== Improved PNO domain selection in PNO-LCCSD(T)-F12 ==== 
 + 
 +A new option **''SORTPNO''** has been added to the PNO program (see [[local_correlation_methods_with_pair_natural_orbitals_pnos#Options for PAO/OSV/PNO generation]] for details). If this is given, first all PNOs that satisfy the occupation number criterion are added. Then further PNOs are added until the energy completeness criterion is also satisfied. With ''SORTPAO'' these PNOs are added in the order of decreasing semi-canonical energy contributions (rather than in the order of decreasing occupation numbers). In some cases this leads to a drastic reduction of some domains. For backward compatibility, please set PNO option ''SORTPNO=0''. Additionally, there a a new option **''EPNO_MIN=//value//''**. PNOs which have semi-canonical energy contributions smaller than the given value are omitted. The default value is zero. 
 + 
 +==== Additional options in the PNO region program ==== 
 + 
 +Additional options to define region-specific domain options and to extend the center list for each region are available, see 
 +[[local_correlation_methods_with_pair_natural_orbitals_pnos#Correlation regions in the PNO program [REGION]]]. 
 +==== Restarts with INTERACT ==== 
 + 
 +The restart possibilities in the INTERACT program have been improved. It is now possible to carry out calculations without counterpoise corrections or monomer relaxation calculations with a restart after a completed CP calculation. See [[properties_and_expectation_values#intermolecular_interaction_energies| Interact]] for details. 
 + 
 +==== CC2 properties and transition moments  ==== 
 + 
 +CC2 Transition moments and ground state density matrices/properties can now be computed using the canonical CC2 program, see section [[the_closed-shell_density_fitting_cc2_program_for_ground_and_excited_states|The closed-shell density fitting CC2 program for ground and excited states]]. 
 + 
 +==== Finite field property calculations ==== 
 + 
 +New automatic procedures to simplify the calculation of properties like multipole moments or polarizabilities using finite field approximations have been added. 
 + 
 +==== Self-consistent random phase approximation method ==== 
 + 
 +The self-consistent random phase approximation method ([[https://doi.org/10.1063/1.4818984|J. Chem. Phys.]] 139, 084113 (2013), [[https://doi.org/10.1103/PhysRevLett.134.016402|Phys. Rev. Lett.]] 134, 016402 (2025)) is now available within the [[Kohn-Sham random-phase approximation#SCRPA program|SCRPA program]]. The program uses the optimized effective potential method to construct the exact exchange potential and the random phase approximation correlation potential. The [[Kohn-Sham random-phase approximation#SCEXX program|SCEXX program]] offers the self-consistent exact exchange optimized effective potential method ([[https://doi.org/10.1063/1.2751159|J. Chem. Phys.]] 127, 054102 (2007), [[https://aip.scitation.org/doi/full/10.1063/5.0056431|J. Chem. Phys.]] 155, 054109 (2021)) that neglects correlation. 
 + 
 +==== Enhanced interface with other codes ==== 
 + 
 +Molpro can now read and write geometries to/from [[https://trex-coe.github.io/trexio/|TREXIO]] files. 
 + 
 +New features have also been added to [[https://github.com/molpro/pymolpro|pymolpro]]: 
 +  * Export geometry, basis set and orbitals to a [[https://trex-coe.github.io/trexio/|TREXIO]] file. 
 +  * [[https://molpro.github.io/pymolpro/pymolpro.html#pymolpro.ASEMolpro|Prepare and launch Molpro runs]] within the [[https://wiki.fysik.dtu.dk/ase/|Atomic Simulation Environment (ASE)]] 
 + 
 +=====  GUI gmolpro 2.9.0 ===== 
 + 
 +gmolpro version 2.9.0 should be used together with Molpro2025.3. On Mac, the package you download is already bundled with this Molpro version. 
 + 
 +=====  GUI gmolpro 2.8.0 ===== 
 + 
 +gmolpro version 2.8.0 should be used together with Molpro2025.2. On Mac, the package you download is already bundled with this Molpro version. 
 + 
 +The basis set library of Molpro had been extended, and gmolpro has been correspondingly adjusted.  
 + 
 +=====  GUI gmolpro 2.7.0 ===== 
 + 
 +gmolpro version 2.7.0 should be used together with Molpro2025.1. On Mac, the package you download is already bundled with this Molpro version. On Linux, gmolpro does not come bundled with Molpro, and you need a separate installation of Molpro.
  
 A trajectory can be visualised, i.e. a set of xyz files appended to each other. A trajectory can be visualised, i.e. a set of xyz files appended to each other.
 +In the builder window, do ''Geometry/Import to builder window...''
 +
 +The box in which orbitals are plotted can be enlarged so that very diffuse orbitals can be plotted.
 +
 +Molpro versions beyond 2025.1 will need gmolpro version at least 2.7.0 because the xml file will be split in xml and xml-sidecar
 +
 +==== Bug fixes ====
 +
 +A bug has been fixed where double hybrid DFT calculations gave different results in serial and parallel calculations. The scaling factor to the MP2 correlation energy is now applied both in serial and in parallel mode (before it was only applied in serial DH calculations). See section [[the_density_functional_program#Double-hybrid functionals (DH, DSDH)|Double-hybrid functionals (DH, DSDH)]] for more info.
 +
 +
 +
 +
 +
 +
 +===== New features of MOLPRO2025.2 =====
 +==== Interact ====
 +
 +A minor bug concerning the PNO-LCCSD(T*) triples scaling has been fixed. The effect of the bug is negligible.
 +
 +==== Fitting basis sets ====
 +
 +The def2-QZVPP MP2FIT basis for Ba has been replaced by a newer version. The old version can be accessed as def2-QZVPPO.
 +
 +==== Nuclear-Electronic Orbital DFT ====
 +
 +Density-fitted NEO-DFT is now available. The program is currently limited to the use of standard numerical integration grids (SG1-3). The electron-proton correlation functionals currently available through the libxc library are the epc17.1, epc17.2, epc18.1 and epc18.2. Multicomponent basis sets (cc-pV//n//Z-mc, aug-cc-pV//n//Z-mc) from Brorsen and coworkers are now provided in the Molpro basis set library.
 +
 ===== New features of MOLPRO2025.1 ===== ===== New features of MOLPRO2025.1 =====
  
 ====  Bug fixes ==== ====  Bug fixes ====
-Various bug fixes. Most notably, a bug that lead to a wrong Hartree-Fock starting guess in the presence of ECPs has been fixed. In some cases an error occurred, in others no or very erratic HF convenience resulted. This only occurred if: (i) a segmented or core-valence (CV) basis set was used; (ii) two atoms of the same kind followed each other in the geometry before an atom with ECP; (iii) and/or symmetry was used which reflected atoms into each other. The error did not occur with [a]vnz[-f12] basis sets.+Various bug fixes. Most notably, a bug that led to a wrong Hartree-Fock starting guess in the presence of ECPs has been fixed. In some cases an error occurred, in others no or very erratic HF convergence resulted. This only occurred if: (i) a segmented or core-valence (CV) basis set was used; (ii) two atoms of the same kind followed each other in the geometry before an atom with ECP; (iii) and/or symmetry was used which reflected atoms into each other. The error did not occur with [a]vnz[-f12] basis sets. 
 + 
 +====  Basis sets ==== 
 + 
 +New Pseudopotentials and associated basis sets have been added for several 3d transition metals.
  
 ====  Interact ==== ====  Interact ====
-A new program to automatically compute CP-corrected interaction energies has been added, see [[properties_and_expectation_values| Interact]]+A new program to automatically compute CP-corrected interaction energies has been added, see [[properties_and_expectation_values#intermolecular_interaction_energies| Interact]]
  
 ====  PNO regions  ==== ====  PNO regions  ====
 The PNO correlation region method [[Local correlation methods with pair natural orbitals (PNOs)#Correlation regions in the PNO program [REGION]|(see REGION)]] has been further improved. The PNO correlation region method [[Local correlation methods with pair natural orbitals (PNOs)#Correlation regions in the PNO program [REGION]|(see REGION)]] has been further improved.
 +
 +==== Binaries and parallelism ====
 +We recently took the decision not to distribute sockets binaries any more. MPI-PR is the configuration [[https://globalarrays.readthedocs.io/en/latest/Section1-BuildInstructions.html#selecting-the-underlying-one-sided-runtime|recommended]] by the developers of Global Arrays, and, as far as we understand, is the only one that is asserted to be supported.  See [[running_molpro_on_parallel_computers|Running Molpro on parallel computers]] for additional information on how configuring memory for Molpro can be slightly simpler when using MPI-PR.  We are aware of some failures on large memory-intensive jobs using runtimes other than MPI-PR, and feel that the correct thing to do is to make available only binaries that use MPI-PR.
  
 ===== New features of MOLPRO2024.4 ===== ===== New features of MOLPRO2024.4 =====
Line 289: Line 392:
   * A new variant of the SO-SCI optimization is available for the RHF case. It optimizes all AVAS orbitals quadratically and is called by ''{RHF,SO-SCI-ACTIVE}''.   * A new variant of the SO-SCI optimization is available for the RHF case. It optimizes all AVAS orbitals quadratically and is called by ''{RHF,SO-SCI-ACTIVE}''.
  
-The ''SO-SCI'' option significantly improves convergence and is recommended in difficult cases. The cost may even be lower than for standard ROHF or CAHF calculations, since the number of iterations is reduced. Furthermore, the probability of convergence to local minima or saddle points is much reduced.+The ''SO-SCI'' option significantly improves convergence and is recommended in difficult cases. The cost may even be lower than for standard open-shell RHF or CAHF calculations, since the number of iterations is reduced. Furthermore, the probability of convergence to local minima or saddle points is much reduced.
    
 See section [[the_scf_program#options_for_the_so-sci_optimization|options for the SO-SCI optimization]] for more details and options. See section [[the_scf_program#options_for_the_so-sci_optimization|options for the SO-SCI optimization]] for more details and options.
Line 366: Line 469:
   * The old version can still be used by specifying option "old", e.g. hf,old.   * The old version can still be used by specifying option "old", e.g. hf,old.
  
-  * Optionally, the hybrid second-order/super-ci method (so-sci) can be used in difficult cases (hf,so-sci). In open-shell cases this converges more robustly than the standard ROHF procedure.+  * Optionally, the hybrid second-order/super-ci method (so-sci) can be used in difficult cases (hf,so-sci). In open-shell cases this converges more robustly than the standard open-shell RHF procedure.
  
 ==== MCSCF/CASSCF ==== ==== MCSCF/CASSCF ====