[CP2K-user] [CP2K:18795] Re: A error in the process of restarting meta-dynamic

[Marcella Iannuzzi]

Hi

I have to insist in remarking that changing the collective variable space 
along a run is conceptually wrong
Even if it were possible to run, whatever you get is not a reliable result. 
The metdynamic potential contributions (the hills) have the dimension of 
the collective variable space, you cannot change abruptly this dimension. 
Also changing the definition of one collective variable along one 
simulation is wrong. 

This said, it is expected that the simulation cannot run, because the 
restarting procedure works by reading all the  past information on the 
spawned hills. If N CV are defined it expects to find these information in 
N dimensions within the restart file, and obviously these are not 
available. 

Regards
Marcella
 

On Wednesday, May 10, 2023 at 1:25:59 PM UTC+2 廉永健（lianyongjian） wrote:

> Dear DR. Marcella
>    Thank you for your response. The detailed operation is as follows: this 
> task has been running for 26 ps, but at 25.5 ps, I found that an unexpected 
> reaction occurred. Therefore, I copied cp2k-1. ener, cp2k. inp, cp2k 
> COLVAR. metadynLog, cp2k HILLS. metadynLog, cp2k-1_ 25500. start, mix. cif 
> to a new folder, and then save the results in cp2k-1. ener, cp2k. inp, cp2k 
> COLVAR. metadynLog, cp2k HILLS. metadynLog to only 25.5 ps, followed by 
> cp2k. inp and cp2k-1_ A new colvar and its corresponding COLLECTIVE section 
> were added to 25500. restart, and an error occurred in the end. Thank you 
> for your help!
> 在2023年5月10日星期三 UTC+8 15:32:17<Marcella Iannuzzi> 写道：
>
>> Dear ?
>>
>> Without additional information it is hard to guess what the problem is. 
>> Apparently the program expects to read data from the restart file, that 
>> are instead missing.
>> Anyway changing the type and number of collective variables along a 
>> metadynamics run is not correct in principle.
>> Regards
>> Marcella
>>
>>
>> On Tuesday, May 9, 2023 at 3:38:29 PM UTC+2 廉永健（lianyongjian） wrote:
>>
>>> Dear all
>>> Recently， I am trying to study the potential energy surface of reaction 
>>> by meta-dynamic. When I add a new colvar and corresponding COLLECTIVE 
>>> section to input file and restart file, and run the restart calculation, 
>>> the program reports an error with the error content being 'tried to go past 
>>> end'. Why? How should I solve this error. Thans for your time and patience.
>>>
>>>

-- 
You received this message because you are subscribed to the Google Groups "cp2k" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cp2k+unsubscribe at googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/cp2k/9b0405f0-9cd3-481f-ad0a-7f549d907fa9n%40googlegroups.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20230510/e2329d0c/attachment.htm>