error while using dymmatrix.pl

[prasadmatukumilli]

error is:
Illegal division by zero at /../dymmatrix.pl line 166
line 166 is:
$matrix->[$current_displacement][$this_displacement] /= (sqrt($mass[$k]*$mass[$i]));

as I dont know perl scripting, pls anyone comment on why $mass[] is coming as zero.?
Thanks in advance.

prasad

[prasadmatukumilli]

for testing here i attached OUTCAR and DISPLACECAR

[graeme]

It doesn't look like you have done a dynamical matrix calculation with our vtstcode. If you do this, the dynmatrix.pl can be used to extract the Hessian and determine the modes and frequencies.

[prasadmatukumilli]

Dear Dr.Graeme,
Now, after dynamical matrix calculation by modifying INCAR with
ICHAIN=1(my objective is to generate movies for each normal mode),
NSW=49(16 atoms in unitcell),
POTIM=0.0,
IBRION=3 (keeping remaining tags as it is from relaxation run)
and its a single image calculation,

while I am running dymmatrix.pl (DISPLACECAR generated by dymselsph.pl with POSCAR 1 8 0.001) (OUTCAR from dynamical matrix calculation)
result giving all frequencies as ZEROS and imaginary.

I have tried changing EDIFFG from -1E-05 to -1E-8 and also to -1E-7,
also tried ISYM = 0 while EDIFFG =-1E-5
and in all the trails, the same result:all zero frequencies.

Please suggest to solve this problem.
The attached outcar is of EDIFFG=-1E-8

[graeme]

You do not have our vtstcode linked into your vasp binary.

[prasadmatukumilli]

Since I cannot link vtstcode to vasp binay right now,
as I have nice results of IBRION=5 run (DYNMAT file, OUTCAR ..)
can you please suggest if there is any way to generate normal mode movie files (or to create modes.dat) by using vtst scripts?

[graeme]

We have not linked up to the IBRION=5 scheme. It should not be difficult, but we have not done it.

[prasadmatukumilli]

please correct me in understanding format of "modes.dat"

is tat a eigen vector matrix ?
x11 y11 z11 x12 y12 z12 ... x1natom y1natom z1natom
x21 y21 z21 x22 y22 z22 ... x2natom y2natom z2natom
. . . . . . . . . . . .
. . . . . . . . . . . .
xndof1 yndof1 zndof1 xndof2 yndof2 zndof2 ... xndofnatom yndofnatom zndofnatom

if this is correct, i can manually create this file from the eigenvectors of ibrion=5 outcar.

[graeme]

Yes, that's right. If you write a script to do it, we can add it.