Hi Dave,
> Note that the total energy is quite close in "good" and "bad" outputs, its
> the heat of formation that is messed up. As a result of all of this, the
> geometry is somewhat different, and the charges slightly different. But
> this
On Cygwin g77 I am getting the correct answer out of mopac. This of course
means that the antechamber test cases fail since the saved files reflect the
results from the "incorrect" mopac runs. E.g.
Sustiva.mol2
1 C1 0.7280 1.4030 0.2550 ca 1 SUS -0.046000
2 H1 -0.1360 1.7560 0.7820 ha 1 SUS 0.165800
3 C2 0.8030 0.0800 -0.1500 ca 1 SUS -0.180700
4 C3 -0.2900 -0.9320 0.1580 c3 1 SUS 0.321300
5 C4 -1.6350 -0.3520 0.0320 c1 1 SUS -0.189400
6 C5 -2.7180 0.1230 -0.0620 c1 1 SUS 0.015400
7 C6 -4.0440 0.6950 -0.1770 cx 1 SUS -0.082100
8 H2 -4.7410 0.3350 0.5580 hc 1 SUS 0.105000
9 C7 -4.6000 1.0130 -1.5410 cx 1 SUS -0.106500
Sustiva.mol2.save
1 C1 0.7280 1.4030 0.2550 ca 1 SUS -0.046500
2 H1 -0.1360 1.7560 0.7820 ha 1 SUS 0.167000
3 C2 0.8030 0.0800 -0.1500 ca 1 SUS -0.174000
4 C3 -0.2900 -0.9320 0.1580 c3 1 SUS 0.328900
5 C4 -1.6350 -0.3520 0.0320 c1 1 SUS -0.189100
6 C5 -2.7180 0.1230 -0.0620 c1 1 SUS 0.016000
7 C6 -4.0440 0.6950 -0.1770 cx 1 SUS -0.084700
8 H2 -4.7410 0.3350 0.5580 hc 1 SUS 0.107000
9 C7 -4.6000 1.0130 -1.5410 cx 1 SUS -0.104600
These look like big differences in the charges to me. Big enough to make
uncomfortable at least. I imagine there are also cases where the differences
are much worse. For example the sustiva test case has a fairly well
optimized structure to begin with.
I think the big issue here though is the fact that the "saved" test case
results are based on the "bad" outputs. This should never have happened
since the results from test cases should be carefully checked by hand before
committing them no?
> is not necessarily a show-stopper for a tutorial/workshop session.
Yes and no. I can see exactly what will happen right now. People will run
this they will check the mopac.out file using due diligence as I instruct
them to do, they will notice that things are wrong and then there will be
tons of questions and I will have to try and explain it. All in all a mess
for a basic workshop.
All the best
Ross
/\
\/
|\oss Walker
| Assistant Research Professor |
| San Diego Supercomputer Center |
| Tel: +1 858 822 0854 | EMail:- ross.rosswalker.co.uk |
|
http://www.rosswalker.co.uk | PGP Key available on request |
Note: Electronic Mail is not secure, has no guarantee of delivery, may not
be read every day, and should not be used for urgent or sensitive issues.
Received on Sun Jul 13 2008 - 06:07:48 PDT
