[CPMD-list] PP for hydrogen
Petr Kulhánek
kulhanek at chemi.muni.cz
Mon Oct 28 20:57:23 CET 2002
Hi,
FHI98PP (release 13-10-1998) contains several bugs.
List of known bugs that author Martin Fuchs provided me in March, 2002 is at the end. In my case, they solved the same problem as yours. If you still have problem with generation of hydrogen PP it will be better when you contact author Martin Fuchs (fuchs at fhi-berlin.mpg.de)
Best regards,
Petr Kulhanek
----------------------------- after release of fhi98PP -----------------------------------------
023 defrtm [defrtm.f] line 3 !* dimension r() to treat elements with Z > 80
...
integer r(4,100)
...
024 defrtm [defrtm.f] line 5 !* allow for pseudo hydrogen with Z < 1
...
if(n .lt. l+1 .or. l .lt. 0) !* remove condition on z
1 stop 'defrtm - illegal argument'
if(z .lt. 1) z = 1 !* add line
...
025 main [fhipp.f] line 295 !* allow for pseudo hydrogen with Z < 1
...
call labelmap(int(z),symz,iexc,symxc) !* replace by
call labelmap(max(int(z),1),symz,iexc,symxc)
...
026 outcore [outwfct.f] line 15 !* correct format
...
700 format(i4,1x,e20.14,2(1x,e20.14)) !* replace by
700 format(e20.14,3(1x,e20.14))
...
027 dsimpson [derlkb.f] line 155 !* initialize variable ie
...
implicit real*8 (a-h,o-z)
include 'parameter.h' !* insert line
...
028 dSSUM [derlkb.f] line 183 !* initialize function dSSUM
...
dSSUM=0.d0 !* insert line
IF(N.LE.0)RETURN
...
029 plgr1 [laguerre.f] line 66 !* avoid possible overflow in factorials
...
parameter(nkmx=34) !* use instead of nkmx=200
...
030 plgr2 [laguerre.f] line 107 !* avoid possible overflow in factorials
...
parameter(nkmx=34) !* use instead of nkmx=200
...
031 klbyii [klbyiiLAPACK.f] line 148 !* initialize variable info
...
info=0 !* insert line
call dspevx('N','I','L',nmx,hessl,-1.d4,1.d1,1,4,1.d-6,
& i,he,hz,nmx,rwk1,iwk1,iwk2,info)
...
032 klbyii [klbyiiLAPACK.f] line 160 !* eliminate variable zero
...
unl(j,i)=0.0 !* use instead of unl(j,i)=zero
...
033 ncpp [ncpp.f] line 398 !* avoid division by zero
...
sgrad=0.16*abs(dcp(i)/pi4)/dc(i)/pi4)**1.33 !* replace by
sgrad=0.16*abs(dcp(i)/pi4)/(max(dc(i),1d-12)/pi4)**1.33
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://cpmd.org/pipermail/cpmd-list/attachments/20021028/8874f633/attachment.html
More information about the CPMD-list
mailing list