5.8 Do loops

Now you have the idea that one geometry is not enough. Why not compute the whole surface? DO loops make it easy. Here is an example, which computes a whole potential energy surface for $\rm H_2O$.

h2o_pes_ccsdt.com

This produces the following table.

 Results for H2O, basis VDZ

     R1    R2   THETA       SCF            CCSD           CCSD(T)
    1.6   1.6   100.0   -75.99757338   -76.20140563   -76.20403920
    1.7   1.6   100.0   -76.00908379   -76.21474489   -76.21747582
    1.7   1.7   100.0   -76.02060127   -76.22812261   -76.23095473
    ...
    2.0   1.9   110.0   -76.01128923   -76.22745359   -76.23081968
    2.0   2.0   110.0   -76.00369171   -76.22185092   -76.22537212

You can use also use DO loops to repeat your input for different methods.

h2o_manymethods.com

This calculation produces the following table.

 Results for H2O, basis DZ, R=1 Ang, Theta=104 degree

 METHOD        E             E-ESCF       E-EFCI
 HF        -75.99897339     .00000000    .13712077
 FCI       -76.13609416    -.13712077    .00000000
 CI        -76.12844693    -.12947355    .00764722
 CEPA(0)   -76.13490643    -.13593304    .00118773
 CEPA(1)   -76.13304720    -.13407381    .00304696
 CEPA(2)   -76.13431548    -.13534209    .00177868
 CEPA(3)   -76.13179688    -.13282349    .00429728
 MP2       -76.12767140    -.12869801    .00842276
 MP3       -76.12839400    -.12942062    .00770015
 MP4       -76.13487266    -.13589927    .00122149
 QCI       -76.13461684    -.13564345    .00147732
 CCSD      -76.13431854    -.13534515    .00177561
 BCCD      -76.13410586    -.13513247    .00198830
 QCI(T)    -76.13555640    -.13658301    .00053776
 CCSD(T)   -76.13546225    -.13648886    .00063191
 BCCD(T)   -76.13546100    -.13648762    .00063315
 CASSCF    -76.05876129    -.05978790    .07733286
 MRCI      -76.13311835    -.13414496    .00297580
 ACPF      -76.13463018    -.13565679    .00146398

One can do even more fancy things, like, for instance, using macros, stored as string variables. See example oh_macros.com for a demonstration.


molpro@molpro.net 2018-12-09