more ksum_error_XHF0: IOError: [Errno 2] No such file or directory: 'wannier90.chk.fmt'

[chao yang]

Dear everyone,

      Now I try to do the LaNiO3 example, however, the ksum_error_XHF0 show me that:

Traceback (most recent call last):

  File ".../bin/XHF0.py", line 291, in <module>

    WAN.load_chk(p["path_bin"])

  File ".../bin/WAN90.py", line 51, in load_chk

    fi_chk = open(self.name + ".chk.fmt", "r")

IOError: [Errno 2] No such file or directory: 'wannier90.chk.fmt'

anyone can help me?

Chao

[Uthpala Herath]

Dear Chao, 

Thank you for your interest in our code!

Did you copy the w90chk2chk.x file from the wannier90 installation directory to the DMFTwDFT/bin directory as explained in https://dmftwdft-project.github.io/DMFTwDFT/installation.html#wannier90-library ?

Best,

Uthpala

--
You received this message because you are subscribed to the Google Groups "DMFTwDFT" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dmftwdft+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/67118b66-4f75-4b59-952b-a220990602e4n%40googlegroups.com.

--

Uthpala Herath

PhD Candidate

Department of Physics and Astronomy

West Virginia University

Morgantown, WV 26505

Tel. (304) 216-2535

Email: ukh...@mix.wvu.edu

Website: https://uthpalaherath.github.io/ 

[chao yang]

Dear  Uthpala,

 I copied the  w90chk2chk.x files from wannier90 to the DMFTwDFT bin and it worked. Thank you for your help.

This time I alway encount a problem that seems related to the memory, as the two ksum* files gives me the details below, please fiind it. 

 Could you give me some suggestions?

Best wishes,

Chao 

 1:

$ more ksum_error_XHF0 

/etc/tmi.conf: No such file or directory

/etc/tmi.conf: No such file or directory

...

 2:

$  more ksum_output_XHF0 

cn1533:SCM:2418:b17e83c0: 58 us(58 us):  open_hca: dev open failed for mlx4_0, err=Cannot allocate memory

cn1533:SCM:2418:b17e83c0: 45 us(45 us): cn1510:SCM:3fdf:eaef23c0: 51 us(51 us):  open_hca: dev open failed for mlx4_0, err=Cannot allocate memory

 open_hca: dev open failed for mlx4_0, err=Cannot allocate memory

cn1532:SCM:1ef6:87d5f3c0: 57 us(57 us): cn1533:CMA:2418:b17e83c0: 31 us(31 us):  open_hca: dev open failed for mlx4_0, err=Cannot allocate memory

 open_hca: getaddr_netdev ERROR:No such  open_hca: dev open failed cn1532:SCM:1ef6:87d5f3c0: 45 us(45 us): cn1533:SCM:2418:b17e83c0: 37 us(37 us):  open_hca: dev open failecn1485:SCM:1b7d:d54633c0: 45 us(45 us

): o open_hca: device mthca0 not found

cn1485:UCM:1d84:263253c0: 34 cn1533:SCM:cn1532:SCM:1ef8:b78ef3c0: 47 open_hca: dev open failed for mlx4_0, err open_hca: getaddr_netdecn1533:SCM:1de6:8d193c0: 43 us(43 us):  open_hca: dev open failed for mlx4_

0, err=cn1533:SCM open_hca: decn1533:UCM:1dd7:d open_hca: device mthca open_hca: dcn1533:SCM:1d99:e2d723c0: 33 us(33 us):  open_hca: devcn1533:SCM:1dd7:dc679cn1533:SCM:1dab:713 open_hca: dev open f open_hca: d

evice ipath0 not found

cn1533:SCM:1d99:e2d723c0: 36 us(36 us):  open_hca: open_hca: dev open failecn1533:SCM:1dab:713063c0: 35 us(35 us):  open_hca: device ipath0 not found

...

[Uthpala Herath]

Hello Chao, 

I'm glad you were able to fix that issue. 

This seems strange and is probably something system related. 

Are you able to test it on a different computer?

Best,

Uthpala

To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/fadb63fd-02cd-491a-89d0-301addd6ac44n%40googlegroups.com.

[chao yang]

Dear  Uthpala,

    I changed the mpi code and fixed the problem.    It's weird that, even do not replaced the mpi code, the DMFT finally goes well after several minutes.
    finally I got the result as some of them shown below the text, please find it.

    Now I have some doubt , the transition metal Ni  possess 8-d electrons and 2-s electrons at the outside-shell, and the oxygen atom have 4-p and 2-s valence electrons , however, the DMFTwDFT_tutorial.pdf told me that, 

  3.4 n-tot

n-tot means the total number of electrons in the Wannier subspace. For exmaple,

in LaNiO3, we have 2 Ni ions with 7 d-electrons/Ni and 6 O ions with 6 p-electrons/O, therefore totally 50 electrons. 

I'm very confusion with the controversy,  may you give me some suggestions?

Best wishes,

yours,

Chao

$ more INFO_ITER 

  # # DMFT  Nd_latt   Nd_imp (Sigoo-Vdc)_latt  (Sigoo-Vdc)_imp  TOT_E(Tr(SigG))  TOT_E(EPOT_imp)

  1   1     8.172057     7.154097       1.333384       1.599140       -87.504257       -62.719405         0.000000

  1   2     8.075553     7.398010       1.374236       1.883057       -78.440456       -60.906406         0.000000

  1   3     8.032673     7.514375       1.416464       1.834478       -73.000080       -60.470733         0.000000

  1   4     7.997161     7.601737       1.452986       1.781975       -68.957698       -60.172757         0.000000

  1   5     7.966920     7.666570       1.487102       1.737852       -65.925217       -59.955245         0.000000

  1   6     7.939823     7.715577       1.516582       1.703163       -63.653130       -59.781478         0.000000

  1   7     7.916296     7.752106       1.543878       1.676342       -61.965203       -59.645471         0.000000

  1   8     7.895055     7.777015       1.567215       1.659033       -60.714840       -59.530392         0.000000

  1   9     7.876416     7.795489       1.586782       1.645471       -59.814400       -59.431190         0.000000

  1  10     7.860317     7.807549       1.603991       1.636657       -59.179435       -59.354696         0.000000

  1  11     7.846537     7.814525       1.619489       1.631893       -58.752343       -59.289052         0.000000

  1  12     7.834423     7.817401       1.631283       1.630103       -58.480059       -59.226033         0.000000

  1  13     7.824321     7.819153       1.641093       1.628941       -58.329853       -59.183699         0.000000

  1  14     7.815906     7.818546       1.649845       1.629292       -58.260351       -59.141945         0.000000

  1  15     7.808712     7.816556       1.656473       1.631357       -58.259985       -59.110279         0.000000

[chao yang]

Dear  Uthpala,

     I got the anwser of the number of electron problem.  Because the Ni lost three electrons therefore Ni have only 7-valence electrons, and the Oxgen atom get the elelctrons from the Ni and La atoms thus have 6-p electrons.   

     It's mine's careless, sorry of about troubling you.

Best wishes,

yours,

Chao 

[Uthpala Herath]

Dear Chao, 

That is correct. You have to take into account their oxidation numbers to calculate the number of electrons in the wannier sub-space. 

To confirm if the number is accurate and that you have selected a good wannier energy window, you can use the electron_count.py script in the DMFTwDFT/scripts directory. 

You just need the same input files you use for the DMFT calculation and then you can run:

$ electron_count.py -np <number of processors> 

This will give you an approximate value of the number of electrons in the wannier subspace through the integration of DOS. 

Thanks, 

Best,

Uthpala

To view this discussion on the web visit https://groups.google.com/d/msgid/dmftwdft/d3cb6d87-4814-4a66-bac8-430f4f4a7ca4n%40googlegroups.com.