About linearly dependent basis in Molpro

[Ioannis Kerkines]

Dear all,

We have encountered this problem a number of times in the present and
older versions of Molpro. I also did not quite understand when this error
happens, but we have noticed something common in all cases: When Molpro
prints in the output the basis functions it will use from its basis set 
library, the basis set entries for an atom are printed twice, although 
in the input it was specified only once.

For instance, the input may read:

basis={
spdf,c,vtz;c;
spdf,o,vtz;c;
}

still the output prints something like:

 Library entry C    S VTZ                  selected for orbital group  1
 Library entry C    P VTZ                  selected for orbital group  1
 Library entry C    D VTZ                  selected for orbital group  1
 Library entry C    F VTZ                  selected for orbital group  1
 Library entry O    S VTZ                  selected for orbital group  2
 Library entry O    S VTZ                  selected for orbital group  2
 Library entry O    P VTZ                  selected for orbital group  2
 Library entry O    P VTZ                  selected for orbital group  2
 Library entry O    D VTZ                  selected for orbital group  2
 Library entry O    D VTZ                  selected for orbital group  2
 Library entry O    F VTZ                  selected for orbital group  2
 Library entry O    F VTZ                  selected for orbital group  2

and then comes the error message about linear dependency. To come around
this problem in this case one has to omit the line

spdf,o,vtz;c;

from the input. Then Molpro will use for O the same basis set as C
(cc-pVTZ and not cc-pVDZ which is the default).

Has anybody else noticed a similar thing in their outputs?

Best regards,
Ioannis