ip rpa eps and eps0

[zhukov_vladlen]

Dear all,
I have just compilated yambo 3.4.1 and done IP and RPA calculations for optics in metallic titanium and rutile. The commands used are usual:
yambo -o c -k hartree,
with Chimod = IP or Hartree in the yambo.in.

What is strange I find in the calculated o.eps...... files of RPA: here the eps-results are almost identical to eps0, with differences no more that 0.00001. I find this for both Re and Im parts, for all the Q-vectors. Moreover, the results of IP-calculos are almost identical to those of RPA. Is this true?

Vladlen Zhukov.

[Daniele Varsano]

Dear Vladlen,
please fill your signature with your affiliation, this is a rule of the forum.
Indeed your results are quite strange, in order to look what is going wrong is always useful to post the output and report files.
In any case I wonder if you raised the variable NGsBlk until convergences, this variable controls the G vector block of the response function taking account of the local field effects. If you leave the NGsBlk=1 no local field are included and the Hartree approximation is equivalent to the IP approximations.
In the tutorial section (fantastic dimension) there are some examples on RPA calculations:

http://www.yambo-code.org/theory/docs/doc_Xd.php#RPA
http://www.yambo-code.org/tutorials/fan ... ns/LiF.php

Best,

Daniele