[molpro-user] saving dump record and using it as initial guess
elkeshavarz5 at sci.ui.ac.ir
elkeshavarz5 at sci.ui.ac.ir
Mon Nov 4 03:54:45 GMT 2013
Dear Daniel
Thanks for your comments.
Again the error is there. It is looking for dump record 2101.2, and it
is there (as it has been listed at the end of previous calculation,
but it is ccsd wave function.)
Regards
Elham
Quoting Dan Kats <dnkats at gmail.com>:
> Dear Elham
>
> You can try to force Molpro to look for 2100.2 instead of 2101.2:
>
> basis=vdz
> geometry={be1;be2,be1,r }
> r=1.6 ang
> {hf;save,2000.2}
> {ccsd
> save,5000.2
> }
> r=1.7 ang
> {hf;start,2000.2}
> {ccsd
> start,5000.2
> }
>
> Regarding your second concern: the only possibility to restart a
> calculation is to use wfu files (ie, by putting "file,2,be2.wfu" in
> your input). You can find the wfu file after a successful calculation
> in your ~/wfu directory, or somewhere in the $TMPDIR directory
> otherwise.
>
> Best wishes
> Daniel
>
> 2013/10/29 <elkeshavarz5 at sci.ui.ac.ir>:
>> Dear Daniel
>>
>> By using the input you sent me, the result is the following error at the
>> start of RHF calculations for the second geometry:
>> ...
>> **********************************************************************************************************************************
>> DATASETS * FILE NREC LENGTH (MB) RECORD NAMES
>> 1 18 7.56 500 700 610 900
>> 950 970 1000 129 960 1100
>> VAR GEOM BASINP SYMINP
>> ZMAT AOBASIS BASIS P2S ABASIS S
>> 1400 1410 1200 1210
>> 1080 1600 1650 1700
>> T V H0 H01
>> AOSYM SMH MOLCAS OPER
>>
>> 2 9 1.35 500 610 700 1000
>> 520 2100 1001 2101 5000
>> VAR BASINP GEOM BASIS
>> MCVARS RHF BASIS CCSD CCSD
>>
>> PROGRAMS * TOTAL INT CCSD HF INT RESTART
>> CPU TIMES * 0.33 0.10 0.03 0.01 0.10 0.00
>> REAL TIME * 0.39 SEC
>> DISK USED * 56.21 MB
>> SF USED * 0.31 MB
>> GA USED * 0.00 MB (max) 0.00 MB (current)
>>
>> **********************************************************************************************************************************
>>
>> 1PROGRAM * RHF-SCF (CLOSED SHELL) Authors: W. Meyer, H.-J. Werner
>>
>>
>> ?ERROR IN GET_INFO: NO DUMP RECORD: 2101.2
>>
>> On the other hand, if I want to do the calculations at several geometries in
>> one input file, if it does not converge at one geometry and I decide to skip
>> that geometry, I have to redo the calculation for the last converged
>> geometry again in order to use the result for the next geometry. Is there
>> any way for escaping from this repeating?
>>
>> Regards
>> Elham
>>
>>
>> Quoting Dan Kats <dnkats at gmail.com>:
>>
>>> Dear Elham
>>>
>>> You can restart your calculation in the same input, i.e., you don't
>>> need a wfu file for it. Your input can look like:
>>>
>>> basis=vdz
>>> geometry={be1;be2,be1,r }
>>> r=1.6 ang
>>> hf
>>> {ccsd
>>> save,5000.2
>>> }
>>> r=1.7 ang
>>> hf
>>> {ccsd
>>> start,5000.2
>>> }
>>>
>>> Best wishes
>>> Daniel
>>>
>>> 2013/10/20 <elkeshavarz5 at sci.ui.ac.ir>:
>>>>
>>>> Dear users
>>>>
>>>> I have convergence problem in a ccsd calculation and I have tried
>>>> different
>>>> options suggested in the manual. Now I want to use the wf of a converged
>>>> calculation for another geometry as initial guess. As it is said in the
>>>> manual, I have printed
>>>>
>>>> file,2,Be2.wfu
>>>>
>>>> at the beginning of the input file. It proceeds without any error, but I
>>>> can
>>>> not find the wfu file!
>>>> Furthermore, if I find it can I use it the same as a dump record as the
>>>> initial guess in another computation? like
>>>>
>>>> ccsd
>>>> start,Be2.wfu
>>>>
>>>> I appreciate any comment.
>>>> Elham Keshavarz
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> ----------------------------------------------------------------
>>>> University of Isfahan (http://www.ui.ac.ir)
>>>>
>>>> _______________________________________________
>>>> Molpro-user mailing list
>>>> Molpro-user at molpro.net
>>>> http://www.molpro.net/mailman/listinfo/molpro-user
>>>
>>>
>>>
>>>
>>> --
>>> Daniel Kats
>>> Institute for Theoretical Chemistry
>>> University of Stuttgart
>>> Pfaffenwaldring 55
>>> D-70569 Stuttgart, Germany
>>> _______________________________________________
>>> Molpro-user mailing list
>>> Molpro-user at molpro.net
>>> http://www.molpro.net/mailman/listinfo/molpro-user
>>>
>>
>>
>>
>>
>>
>>
>>
>>
>>
>> ----------------------------------------------------------------
>> University of Isfahan (http://www.ui.ac.ir)
>>
>
>
>
> --
> Daniel Kats
> Institute for Theoretical Chemistry
> University of Stuttgart
> Pfaffenwaldring 55
> D-70569 Stuttgart, Germany
>
----------------------------------------------------------------
University of Isfahan (http://www.ui.ac.ir)
More information about the Molpro-user
mailing list