[molpro-user] read wavefunctions

[binjiang.nju]

Dear molpro developers and users,

    I have been running CASSCF calculations using Molpro 2015.1 and in many cases would like to read the converged wavefunctions from the last calculations as the initial guess. While I recently noticed that the wfu file was sometimes not correctly load leading to errors in the subsequent CASSCF calculation. I am not sure if it is a bug or something wrong of our installation. 

For example, we read the same wfu file that we saved in a prior calculation, we could get this:

 Permanent file  2  c3h4n4o-conical-s1s2-avdz.wfu    assigned.  Implementation=df

without any other information in the wfu file.

otherwise, we could get this:

 PROGRAM * RESTART


 Reading variables from file 2

 _NUMVAR         =    924.00000000
 _DKROLL_DONE    =      0.00000000
 _FOCKDONE       =      1.00000000
 _INTDONE        =      1.00000000
 _SCFDONE        =      1.00000000
 _DIRECT         =      0.00000000
 _DKHO           =     -1.00000000
 _BASIS          =    AVDZ
 _ZUNIT          =    ANG
 _NELEC          =     58.00000000
 _PROGRAM        =    MULTI
 _CPUSTEP        =    375.76000000
 _SYSSTEP        =     40.42000000
 _WALLSTEP       =    417.10000000
 _BASINP         =    610.00000000
 _CPUTOT         =    386.37000000
 _DFTEXFAC       =      1.00000000
 _DMX(1:8)       =     -1.65394570    -0.86670926    -0.71821722    -0.36845989    -0.57203860    -0.83392129
                       -1.05871272    -0.36906175
 _DMY(1:8)       =      0.07947334    -0.33528898     0.83936831     0.90064253     0.10738726    -0.42006850
                        0.19313844     0.96052234
 _DMZ(1:8)       =      0.23188526     0.24157445     0.21044843     0.23376887     0.22569465     0.25206690
...

And we also got the following things in the same set up

 PROGRAM * RESTART


 Reading variables from file 2

 _NUMVAR         =    921.00000000
 _DKROLL_DONE    =      0.00000000
 _INTDONE        =      1.00000000
 _DIRECT         =      0.00000000
 _DKHO           =     -1.00000000
 _BASIS          =    USERDEF
 _ZUNIT          =    ANG
 _NELEC          =     58.00000000
 _PROGRAM        =    INT
 _CPUSTEP        =     23.71000000
 _SYSSTEP        =      5.72000000
 _WALLSTEP       =     29.49000000
 _BASINP         =    610.00000000
 _CPUTOT         =     24.26000000
 _CPUSORT        =     19.76000000
 _SYSSORT        =      4.67000000
 _WALLSORT       =     24.46000000
 _ENERGY_METHOD  =    MULTI
 _ENERGY_BASIS   =    cc-pVDZ
 _ENUC           =    359.07181343
 _SEIG_MIN       =      0.00004764

    The error is so weird that I can not predict when I got the correct results before checking the out file. Can someone help me solve this problem? Thanks in advance.

Best, Bin



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.molpro.net/pipermail/molpro-user/attachments/20161103/5a14957a/attachment.html>