Help: error about optimization

[qingfang]

Dear Sir:
I am not good at using the ATK.And I want to optimize my module,but It always made mistakes,like this:


Traceback (most recent call last):
File "c:\users\admin\appdata\local\temp\6959826860399464.py", line 111, in <module>
trajectory_filename=None,
File ".\zipdir\NL\Optimization\OptimizeGeometry.py", line 57, in OptimizeGeometry
NL.ComputerScienceUtilities.Exceptions.NLValueError: OptimizeGeometry is only supported for LCAO & EMT calculators.
 
In addition,the version which I used  is  the trial version of  2010.8.b1.
Can you tell me the differences between the trial version and the authorised version ?
because sometime when I caculate some quantities  It can be work without error in the authorised version.
So I am grateful for your help!

[kstokbro]

I think you are using the HuckelCalculator, it does not support optimization. Use the LCAOCalculator, i.e. ATKDFT.
I strongly recommend to download the released version, there are many differences and many bugs has been corrected relative to the beta release.

[qingfang]

I didn't save  the trajectory file, and the K point which I set was (1,1,100).
In addition,the optimization information of the script about my module had several words"warning:truncated content".
right now,the caculation had finished. but I can't find the output file.

[kstokbro]

the program sends output to stdout, if nothing is specified this is the screen.
in linux write
atkpython input_file.py > output_file.log

however you should have a .nc file with the relaxed geometry, which can be studied with vnl

[qingfang]

i use the 2010.8 version in windows system. i have set the default output nc file which called "analysis.nc".but i only find the  molecule configuration which was not the   final output file ,but may be the input file.

[Anders Blom]

Are there one or two entries in your NC file? What is the script you ran?

[qingfang]

i have  default output file and  trajectory file ,but with the different name.
and  i ran the  py  script file which i already seted the parameters.
right now ,i  finished my caculation, but i don't know how to find the  model which i optimized.
thanks for your teaching.

[qingfang]

i  can't find the result. in the old version ,there will be a vnl file ,and i can see the module clearly.but in the new version ,there will be not .so ,how can i see the final result .
thank you .

[qingfang]

and can i read the final result  if i didn't save the trajectory file?
thank you very much!

[zh]

The atomic coordinates are usually printed out in the log file during the self-consistent calculations for each step in the geometry optimization. You can extract them to build a file with xyz format.

[Anders Blom]

If you are using the 10.8 (final!) version, you should have in the NC file two molecular configurations, one was input and one is output (the optimized geometry). If not, it will be caused by something in the script. Unless you post the script used, it is very difficult to offer proper assistance. You can remove the geometry definition if you prefer to keep it secret, but keep the rest of the script. If your script was GENERATED by 10.8.b1 and you RUN it with 10.8, the problem will occur that only the input geometry is saved to the NC file. You need to generate the script also with 10.8 for it to work properly. This was one of the known bugs in the beta-release which were fixed in the final version.

[qingfang]

thank you very much.