From NWChem
Viewed 5202 times, With a total of 7 Posts

Clicked A Few Times
Threads 12
Posts 18


4:05:35 PM PDT  Wed, Jul 10th 2013 

Hi,
In dft or hf calculation, are there any command which can turn on printing the overlap matrix of the atomic orbitals?
thanks,




Niri Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop


Forum Regular
Threads 4
Posts 209


4:55:05 PM PDT  Wed, Jul 10th 2013 

Hi Alexx,
Do the following in your input file
scf
print "overlap"
end
task scf
For the kinetic integrals, do print "kinetic"
Best,
Niri
niri.govind@pnl.gov




Niri Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop


Forum Regular
Threads 4
Posts 209


4:55:28 PM PDT  Wed, Jul 10th 2013 

This will print the integrals in your output file.




Niri Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop


Forum Regular
Threads 4
Posts 209


5:10:31 PM PDT  Wed, Jul 10th 2013 

If you want the ao overlap matrix, then
scf
print "ao overlap"
end
task scf




Clicked A Few Times
Threads 12
Posts 18


5:51:59 AM PDT  Thu, Jul 11th 2013 

Hi, Niri
Thanks for your reply.
I tried both.
For
scf
print "ao overlap"
end
It did print out the ao overlap matrix, but the precision is just 5 decimals. Are there any flag that can be turned on to increase the decimal number to 8.
For
scf
print "overlap"
end
It could print out the overlap integrals line by line, and the precision is good, 8 decimals.
But after perusal, I found it misses some overlap matrix elements, compared to the previous situation.
Can you give me some suggestion about this?
Quote:Niri Jul 10th 6:10 pmIf you want the ao overlap matrix, then
scf
print "ao overlap"
end
task scf




Huub Forum:Admin, Forum:Mod, NWChemDeveloper, bureaucrat, sysop


Forum Regular
Threads 1
Posts 185


10:03:57 AM PDT  Thu, Jul 11th 2013 

Hi Alexx,
with
scf
print "overlap"
end
the code will print only the nonzero overlap integrals. So indeed you are right in that not all integrals are printed. But the ones that are not printed are smaller than 1.0e8.
Huub




Clicked A Few Times
Threads 12
Posts 18


12:01:32 PM PDT  Thu, Jul 11th 2013 

Hi, Huub
Thanks for your reply.
I code up a small parser to scan the output from "print "overlap"" line by line to store the overlap integrals into a matrix. Then when I printed it out, I found that the diagonal elements are not all one. Some are nonzero and lessthanone values, which means the original atomic orbitals are not normalized. Some are zeros, which means it misses from output.
For those nonzero and lessthanone values, they can be confirmed by the output from "print "ao overlap"". For those zero values, they are actually ones in the output from "print "ao overlap""
Can you offer some hints about this?
Quote:Huub Jul 11th 11:03 amHi Alexx,
with
scf
print "overlap"
end
the code will print only the nonzero overlap integrals. So indeed you are right in that not all integrals are printed. But the ones that are not printed are smaller than 1.0e8.
Huub




Clicked A Few Times
Threads 12
Posts 18


1:34:10 PM PDT  Fri, Jul 12th 2013 

Hi, Huub
It seems that the "print "overlap"" does not work in dft calculation.
Like this:
dft
print "overlap"
end
This prints out nothing about overlap.
Do you konw are there any ways to turn on the overlap printing line by line in dft calculation?
thanks
Quote:Huub Jul 11th 11:03 amHi Alexx,
with
scf
print "overlap"
end
the code will print only the nonzero overlap integrals. So indeed you are right in that not all integrals are printed. But the ones that are not printed are smaller than 1.0e8.
Huub



AWC's:
2.5.10 MediaWiki  Stand Alone Forum Extension
Forum theme style by: AWC