Firefly and PC GAMESS-related discussion club

Learn how to ask questions correctly

Vyacheslav
kreme_vg@chemy.kolasc.net.ru

Dear Alex,
I have not understood, is occurrence of asterisks result of quasi-linear dependence or of using of the big basis set? That is, at absence of quasi-linear dependence will be these asterisks appear in files instead of numbers? Sometimes I received such outs files with small basis sets too. That is, apparently, it is a problem not only of the big basis sets and linear dependence can arise for small bases too.

The way of basis set correction not seems good to me. It is difficult for me to predict consequences of my correction of basis. Probably, there are more simple ways of the decision of this problem. What could you advise still?

Best regards,
               Vyacheslav

>============================================================
>On Tue Aug 24 '10 6:00pm, Alex Granovsky wrote
>----------------------------------------------
>>Hi,

>>do you mean stars in the orbital printout section?
>>This is not the bug but rather a standard feature of Fortran
>>formatted I/O. The number is too large to fit into allowed
>>decimal template, and so it simply cannot be printed. Instead,
>>Fortran I/O library prints stars to indicate this event.

>>This problem does not seem to have easy solution because if
>>we enlarge the number of printed decimal digits in the orbital
>>section, this can cause various compatibility issues with various
>>visualization programs, most likely even with ChemCraft itself.

>>Anyway, such a large orbital coefficients is the sign of severe
>>quasi-linear dependence of the basis set that generally should
>>be avoided by deleting some basis functions.

>>The orbitals in the punch file are always free of this problem,
>>as the scientific notation is used here for printout. You can try
>>to open punch file provided you find the program that can read
>>punch files and visualize orbitals.

>>Regards,
>>Alex Granovsky