[molpro-user] problems when using the trial version

Dimitri Cavalcanti dimcavalcanti at hotmail.com
Tue Jan 26 14:41:46 GMT 2010


Hello dear molpro users,

    I'm a new user of molpro, and i start using the trial versioon of the program. Naturally i have a few problems to run my jobs,  here are my questions:


1-) Using the GDIRECT keyword the program will try to store the integrals in memory, but when i use this keyword the program used all my 4Gb of ram and my 8Gb of Swap partition and crash (and carry my gnome desktop with it). When i don't use this keyword it uses 78Gb of my HDD with the integrals (using molpro -n 4 job.com), there is any way to reduce the size of interals to put it in memory?

2-) I'm trying to run some geometry optimization of uranium complexes, and i'm finding some problems when i use the Stuttgart ECP's 79MWB, 80MWB and 81MWB (those who are not implementend in the trial version), i run the job and after computing the initials integrals he give this error message


 ? Error
 ? unknown occupation
 ? The problem occurs in atomic_occupation

But, i already define the occupation, and thw wavefunction of without error.

3-) When i use the 60MWB instead the other LC-ECPs, the program starts and run smoothly (without prompting the previous error message - and it only happens when i copy the basis and the ECP of the file basis.bas in the /usr/local/lib/molpro folder), but when i start the opzimitation of geometry the program crash and print this error message:

 Convergence:                0.00000000  (line search)     0.45000031    20.94413622  (total)
 ? Error
 ? No convergence
 ? The problem occurs in rhfpro

I try to change the method of optimization to QSD or DIIS, and i have the same problem.

So, there are my problems, if anyone knows how to fix i'll apreciate the help.

If it is an input problem here it one sample:

***,job.com
Memory,300,m;
Gprint,basis;                ! Print basis information 
Gprint,orbital;                ! Print orbitals in SCF and MCSCF 

set,CHARGE=-2
geometry={ANGSTROM;
7 ! number of atoms 
GeomXYZ
Cl,    1.785000,    1.001000,    -1.637000
Cl,    -0.222000,    -2.151000,    -1.491000
Cl,    -1.907000,    1.201000,    -1.341000
U,    -0.000000,    0.000000,    0.000000
Cl,    1.907000,    -1.201000,    1.341000
Cl,    0.222000,    2.151000,    1.491000
Cl,    -1.785000,    -1.001000,    1.637000
}
basis={
S,Cl,STO-3G
P,Cl,STO-3G
d,Cl,STO-3G
!id=ECP80MWB elementType=U ECP
ecp,U,80,4,0
1
2,1.000000,.000000
2
2,3.411866,101.510029
2,2.367863,-12.398152 
2
2,2.665803,61.870048
2,1.966819,-0.817262
2
2,1.626020,27.680805
2,1.841639,.948957
2
2,1.277544,-5.060991
2,.882381,2.081285 

!  (7s6p5d)/[4s3p3d] for U ECP80MWB (Q=12), Ref 45.
s,U,5.8004295,3.8669530,2.5600380,.53513283,.25349409,.074301637,.029483039
c,1.3,-0.15626825,.93903619,-.013531200
c,4.5,1.0008500,.35032120
p,U,3.2047425,2.1364950,.76545130,.40121642,.19389215,.081988941
c,1.3,0.15408684,-.41709359,.33525218
c,4.5,0.60207835,.24967068
d,U,1.1732236,.78214904,.34496552,.15000000,.056543751
c,1.3,-0.13063236,.19447560,.39054322
}

{rhf;occ,19,39;closed,19,39;wf,116,0,0}
optg
---

Thanks for all the help, Dimitri P Cavalcanti




 		 	   		  
_________________________________________________________________
Deixe seu computador compatível com a sua vida. Clique para conhecer o Windows 7!
http://www.microsoft.com/brasil/windows7/default.html?WT.mc_id=1539
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20100126/d817273f/attachment.html>


More information about the Molpro-user mailing list