[molpro-user] DDR in Geometry Loop

[Matt Thompson]

Folks, I'm hoping one of you has seen this before and can help me.  I
have a rather complex input file + user procedures that calculates 6
low-lying states via CI and then, among other things, does the DDR
(i.e., NACME) between all the states.  I do this following the lif_nacme
example in the manual, 3-point formula.

The weird thing is the results just seem wrong.  I have results from a
similar molecule that I trust which gives around 0.1 as a maximum
absolute value for any result, so I suspect I should be around that
order of magnitude.  Instead I see values like this (over 4 points, 20,
18, 16, 15 bohr, shown in order of calculation in loop):

dr=0.01           dr=0.001          dr=0.0001
-0.0009027736     -0.0009027736     -0.0009008825
0.0051087741      0.0510876905      0.5108769046
0.0186428064      0.1864419167      1.8644191671
0.0314546272      0.3144831469      3.1448314693

As you can see, it seems like the first point is fine, but when the loop
changes the geometry, I get just wrong values; moreover, they seem to be
proportional to dr, which doesn't seem correct.

Now, then, if I do each point in a separate job, I get:

dr=0.01           dr=0.001           dr=0.0001
-0.0009022745     -0.0009027736      -0.0009008825
-0.0031524102     -0.0031520858      -0.0031478115
-0.0110165455     -0.0110162999      -0.0110144216
-0.0203811660     -0.0203811049      -0.0203764129

This seems more like I'd expect.

To try and figure this out, I've tried not doing diabatic.  No diff.
I've also tried various data,truncate,xxxx.x commands (down to 1100.2)
based on an old Molpro94 job, but no luck.  My orbitals are
2140.2-2142.2 and the CI;TRANS densities are 9901.2-9903.2, so maybe I'm
stepping on the toes of some file?  I'm not sure what files DDR is
using.

I might need to submit this to Bugzilla, but I thought I'd try here
first.

Thanks,
Matt Thompson

-- 
Learning just means you were wrong and they were right. - Aram
   Matt Thompson -- http://ucsub.colorado.edu/~thompsma/
   440 UCB, Boulder, CO  80309-0440
   JILA A510, 303-492-4662