[molpro-user] HF convergence

[Benj FitzPatrick]

Diana,
My first thought is that if you have to use hundreds of cycles in SCF and
CCSD to get convergence there's a deeper problem that needs addressing. For
troublesome SCF convergence it is often useful to see if small basis sets
will converge (sto-6g, 6-31g(d), etc.). If those do you can use a procedure
to run a progression of hf calculations with progressively larger basis
sets.

Another option is to use MCSCF with one active orbital to get the SCF
solution and then get canonical HF orbitals for the CCSD calculation. Kirk
Peterson has posts on the mailing list of how to do this.
Thanks,
Benj


On Sun, May 26, 2013 at 9:42 AM, Diana Rodríguez <diradiana at gmail.com>wrote:

> Hi Benj,
>
> I'm enclosing the last lines of .log corresponding to the other files.
> Best regards,
>
> Diana
>
>
> 2013/5/24 Benj FitzPatrick <benjfitz at gmail.com>
>
>> Hi Diana,
>> Could you send me the last few hundred lines of the molpro.log file as
>> well? They should contain a completed hf and ccsd calculation, and the hf
>> calculation that failed.
>> Thanks,
>> Benj
>>
>>
>> On Fri, May 24, 2013 at 3:20 PM, Diana Rodríguez <diradiana at gmail.com>wrote:
>>
>>> Dear Benj,
>>>
>>> here you will find enclosed the input and the output file. Thanks for
>>> the quick answer.
>>> Best regards,
>>>
>>> Diana
>>>
>>>
>>> 2013/5/24 Benj FitzPatrick <benjfitz at gmail.com>
>>>
>>>> Hi Diana,
>>>> Could you please send the input file you used?
>>>> Thanks,
>>>> Benj FitzPatrick
>>>>
>>>>
>>>> On Fri, May 24, 2013 at 9:28 AM, Diana Rodríguez <diradiana at gmail.com>wrote:
>>>>
>>>>> Hi:
>>>>>
>>>>> I am a new user of MOLPRO with the following problem. I´m doing IRC
>>>>> path calculations (CCSD(T) level in system HSO2) and some times
>>>>> no-convergence in RHF calculations stops the calculus. I have tried
>>>>> all the suggestions in the MOLPRO manual, but I can not avoid this error.
>>>>> Any suggestions, please?
>>>>>
>>>>> _______________________________________________
>>>>> Molpro-user mailing list
>>>>> Molpro-user at molpro.net
>>>>> http://www.molpro.net/mailman/listinfo/molpro-user
>>>>>
>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20130526/e603d246/attachment.html>