#8 CP2K and VASP probrems/improvement

[Kirill Okhotnikov]

Dear Torbjörn,

Your program is very useful. I want to submit some bugs a found and solutions for it.

VASP:

It is good to have a customized "prioritylist = ["_d", "_pv", "_sv", "", "_h", "_s"]" in ESPInterfaces.py. My solution is

   try:
      pl = os.environ['VASP_PP_PRIORITY']
      prioritylist = pl.split(",")
    except:
      prioritylist = ["_d", "_pv", "_sv", "", "_h", "_s"]

It looks like than you have almost the same code for "class POTCARFile" and "class INCARFile". It is good to have some refactoring for the classes.

CP2K:

The version of cp2k 2.5.1 complain about UNIT ANGSTROM in cp2k structure file. Anyhow this unit is default so you can easily exclude this from the file. My suggestion is to disable 2 lines in ESPInterfaces.py

filestring += " UNIT ANGSTROM\n"

filestring += " PERIODIC XYZ\n"

[Torbjörn Björkman]

Hi Kirill,

Thanks for the input, I have put in your solution for custom PP priority. There is a bit of duplicate code between INCAR and POTCAR files, yes, it would probably be a good idea to have a separate function for some of that. I'll open a separate ticket for that and see when I get time to do it.

Comparing the manuals for the CP2k trunk and previous versions, it seems that they made some pretty major changes to their interface recently.
The latest manual,
http://manual.cp2k.org/trunk/CP2K_INPUT/FORCE_EVAL/SUBSYS/PRINT/ATOMIC_COORDINATES.html
now seem to want me to put things in curly brackets {}.
Can you tell me if this is correct?

Second question to you as a user: CP2k is known to me as a somewhat volatile code (as demonstrated by the revised input format...). Does it make sense to support older formats as well, or should I try to just update the interface as I get bugreports from users as they switch to the latest version?

Cheers,
Torbjörn

[Kirill Okhotnikov]

Hi, Torbjörn.

Good to hear your reply so soon. I really glad that the changes, I mentioned will be available in the next version. Of course, refactoring of the code is not urgent. The code is good and I can start VASP calculations almost immediately after applying cif2cell. By the way, do you have a plan to include support for different input format, like pdb or MD/QM program output?

For cp2k 2.5 it should be like this, for cell and for atomic coordinates if you want.

 &SUBSYS

   &PRINT 
     &CELL
       UNIT angstrom
     &END CELL
  &END PRINT

  ...  ...

&END SUBSYS

But, to be compatible with older version it is better to exclude this totally. I checked the documentation for version 2.1 also, it optional also there (and default is angstrom).

Best,
Kirill.

[Torbjörn Björkman]

Very good, I'll just scratch the 'UNIT' keyword. It is very easy to add output in bohrradii or nm, if you feel that there may be a need for CP2k users to have these.

I implement new interfaces as I get requests. If you have something that you need and can supply working, non-trivial examples of the format, it usually takes little more than an hour to code it up. The limit is mostly the lack of requests (but also to some extent my lack of hours...). I haven't gone for common chemistry formats, since there is OpenBabel which should (I think) cover that, but if you want something, just tell.

Cheers,
Torbjörn

[Kirill Okhotnikov]

I don't think that cp2k users will need this. But, may be, somebody)))

As for another input format I would like to use DL_POLY output directly for input of your program. I will have more time in the begging of September and I think, I will implement support for this. I will appreciate if you will include this in your program.

Best,
Kirill.

[Torbjörn Björkman]

The way I handle that myself right now is to convert everything to .cif and
then use cif2cell to convert to whatever. There is for example a vasp2cif
program which maybe will be distributed with cif2cell later:
https://github.com/egplar/vasp2cif
That way may be easier than to implement it directly in cif2cell. Writing
cifs is easy, reading them is a lot harder...

2014-08-07 17:18 GMT+03:00 Kirill Okhotnikov okirk@users.sf.net:

I don't think that cp2k users will need this. But, may be, somebody)))

As for another input format I would like to use DL_POLY output directly
for input of your program. I will have more time in the begging of
September and I think, I will implement support for this. I will appreciate
if you will include this in your program.

Best,
Kirill.

---

• [bugs:#8] http://sourceforge.net/p/cif2cell/bugs/8 CP2K and VASP
  probrems/improvement*

Status: open
Group: v1.0_(example)
Created: Thu Aug 07, 2014 08:35 AM UTC by Kirill Okhotnikov
Last Updated: Thu Aug 07, 2014 11:33 AM UTC
Owner: nobody

Dear Torbjörn,

Your program is very useful. I want to submit some bugs a found and
solutions for it.

VASP:

It is good to have a customized "prioritylist = ["_d", "_pv", "_sv", "",
"_h", "_s"]" in ESPInterfaces.py. My solution is

try:
pl = os.environ['VASP_PP_PRIORITY']
prioritylist = pl.split(",")
except:
prioritylist = ["_d", "_pv", "_sv", "", "_h", "_s"]

It looks like than you have almost the same code for "class POTCARFile"
and "class INCARFile". It is good to have some refactoring for the classes.

CP2K:

The version of cp2k 2.5.1 complain about UNIT ANGSTROM in cp2k structure
file. Anyhow this unit is default so you can easily exclude this from the
file. My suggestion is to disable 2 lines in ESPInterfaces.py
filestring += " UNIT ANGSTROM\n" filestring += " PERIODIC XYZ\n"

---

Sent from sourceforge.net because you indicated interest in
https://sourceforge.net/p/cif2cell/bugs/8/

To unsubscribe from further messages, please visit
https://sourceforge.net/auth/subscriptions/

--

Torbjörn Björkman

[Kirill Okhotnikov]

Thank you for the link. You are right. OpenBabel does the job.

Kirill.

[Torbjörn Björkman]

The changes are implemented in version 1.2.1 which I just uploaded. I'll close this ticket now.

Thanks!
Torbjörn