<table cellspacing="0" cellpadding="0" border="0" ><tr><td valign="top" style="font: inherit;"><DIV>Dear all,</DIV>
<DIV> </DIV>
<DIV>I did the following caspt2 optimization,</DIV>
<DIV><BR>***, test</DIV>
<DIV>geometry={<BR>h;<BR>f,1,r1<BR>}</DIV>
<DIV>r1=0.9 angstrom<BR>basis=3-21g</DIV>
<DIV>hf;<BR>{multi;frozen,1,0,0,0;closed,2,0,0,0;occ,4,1,1,0;}<BR>{rs2;core,1,0,0,0;closed,2,0,0,0;occ,4,1,1,0;}<BR>optg;<BR>---</DIV>
<DIV> </DIV>
<DIV>Then there is an error:</DIV>
<DIV> </DIV>
<DIV> Inconsistent occupation in mcscf and rs2</DIV>
<DIV> </DIV>
<DIV> RS2 gradients can only be computed if the orbitals spaces are identical<BR></DIV>
<DIV> iact (rs2): 2 1 1<BR> iact (mcscf): 2 1 1<BR> iclos (rs2): 2 0 0<BR> iclos (mcscf) 1 0 0<BR></DIV>
<DIV> ? Error<BR> ? Inconsistent occupation<BR> ? The problem occurs in check_rs2occ</DIV>
<DIV> </DIV>
<DIV>It seems the problem comes from the frozen core in multi. I also tried frozen,0,0,0,0; and freeze,1.1; in multi, but it doesn't work.</DIV>
<DIV><BR>Is it possible to get caspt2 analytical gradient based on casscf with frozen core? Thanks.</DIV>
<DIV> </DIV>
<DIV>Wenli</DIV></td></tr></table><br>