Analyse excitons

[young]

Dear developers

After finishing the BSE calculation, I want to analyse excitons, so I run ypp -J <dir> -e s. However, I find a problem. If I using the rim_cut(Random Integration Method or truncated coulomb potential), I will get a error ( STOP signal received while in :[06] Excitonic Properties [ERROR] ndb.BSS_diago, SEC 1, not compatible with ypp input). But if I do not use RIM and Cutoff, I will get correct results. So I want to know why that is. What wrong I am? I used YAMBO.4.3.2. Thanks very much!


Best
Ke

Ke Yang
PHD student
Rensselaer polytechnic institute， Troy, NY, US
Hunan university, changsha, Hunan, China

[Daniele Varsano]

Dear Ke Yang,
thanks for reporting, we still need to fix this problem properly, anyway if you temporary move to another directory the ndb.cutoff, ndb.RIM, ndb.QP databases, and let ypp read the ndb.BS* database only, it should work. In any case, please check that QP correction and BSE run were done with the same number of RIM components and G vectors.

Best,
Daniele

[young]

Dear Daniele,

Thanks for you reply quickly. However, I yet do not get any results including the "rim_cut", following what you said in last post(I move these files (ndb.RIM, ndb.QP databases) to another directory.) . I check ndb.BS_diago_Q01 file, I only find there is a different: if including rim_cut in input, ndb.BS_diago_Q01 file shows the W_Cutoff = "box z ; while, it shows the W_Cutoff = "none , without rim_cut.

And I try another test, after calculating with "rim_cut", I keep other files, and only replace the ndb.BS_diago_Q01 with the file generated by calculating without "rim_cut". I find it can work. Although,it is not reasonable, I think it show the bug maybe in ndb.BS_diago_Q01 file with "rim_cut". Attached the files with“Rim_cut”.

In a word, I find if there are "rim_cut", ndb.BS_diago_Q01 will not work. Can you help me how to analyze the excitonic properties with "rim_cut" ? Thanks in advanced!


Best
Ke

Ke Yang
PHD student
Rensselaer polytechnic institute， Troy, NY, US
Hunan university, changsha, Hunan, China

[Daniele Varsano]

Dear Ke,
you can try the following solution:
in src/io/variables_BS.F
1

Code: Select all

89        call io_elemental(ID,VAR="W_Cutoff",VAR_SZ=1,CH0="",MENU=0)
190        i_local_desc=i_local_desc+1
191        call io_elemental(ID,&
192 &           VAR="    |Coulomb Cutoff          :",CH0=cut_geometry,CHECK=.true.,OP=(/"=="/),&

At line 192 change CHECK=.true. in WARN=.true.
in the same file:

Code: Select all

call io_elemental(ID,VAR="W_DbGd",VAR_SZ=2,MENU=0)
212    i_local_desc=i_local_desc+1
213    call io_elemental(ID,I0=RIM_ng,&
214 &     VAR=' RIM|RL components      [col]:',CHECK=.true.,OP=(/"=="/),&
215 &     DESCRIPTION=local_description(i_local_desc),DO_NOT_DUMP=.TRUE.)
216    i_local_desc=i_local_desc+1
217    call io_elemental(ID,I0=RIM_n_rand_pts,&
218 &     VAR='    |Random points      [col]:',CHECK=.true.,OP=(/"=="/),&
219 &     DESCRIPTION=local_description(i_local_desc),DO_NOT_DUMP=.TRUE.)
220    !
221  endif

at lines 214 and 218:
change CHECK=.true. in WARN=.true.

then recompile the code. In this way the code should emit a warning and continue instead of stopping.
Best,
Daniele

[young]

Dear Daniele,

After re-compiling code based on what you said. It work now. Thank you very much

Best
Ke

Ke Yang
PHD student
Rensselaer polytechnic institute， Troy, NY, US
Hunan university, changsha, Hunan, China