basis set format

[lath... at gmail.com]

I am working on a project where I iteratively call a all-electron DFT 
calculation in cp2k and, among other values, take the basis sets. Now, I 
have the problem, that the contraction coefficients in the basis set file 
differ from the coefficients which are used within the atom-routine 
(atom_info(in, im)%atom%basis%cm). Somehow, the coefficients in the basis 
set seem to be normalized since they are unity whenever only one primitive 
GTO is contracted. Unfortunately, I was not able to figure out, how this 
normalization is done.

Example basis file

C DZVP-ALLELECTRON DZVP-ALL
  6
  1  0  0  6  1
     2808.0640000000   0.0020178300
      421.1383000000   0.0154332000
       95.5866200000   0.0755815500
       26.7390000000   0.2478282000
        8.4328270000   0.4793725000
        2.7605820000   0.3338344000
  2  0  0  2  1
        5.4470040000  -0.0778407700
        0.4792422000   0.5689560000
  3  0  0  1  1
        0.1461565000   1.0000000000
 ...

 
Coefficients called by cp2k:
  6
  1  0  0  6  1
     2808.0640000000   1.967...
      421.1383000000   3.625...
       95.5866200000    5.838...
       26.7390000000    7.362...
        8.4328270000     5.993...
        2.7605820000     1.806...
  2  0  0  2  1
        5.4470040000    -0.701...
        0.4792422000     0.828...
  3  0  0  1  1
        0.1461565000     0.597...
 ...


Of course the first guess would be that they are normalized with respect 
the primitive GTOs or the contracted one but neither works. So I would 
really appreciate an idea of somebody how knows something about the cp2k 
basis set format.


Thank you very much


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.cp2k.org/archives/cp2k-user/attachments/20180117/72f310df/attachment.htm>