Bug report

Posted on : September 13, 2016 (Tue) 11:29:46

by DUONG, DINH LOC

Hi Hitoshi,

I have a bug when using the code. When I get the band structure using spc, there is a error: "tmp000", line 6: Matrix does not represent a grid. I am sure this message is from the gnuplot and it seems the output file is not the matrix type.

I have open the file by text and and see some line, there is no space between them. It seems the number of character is larger then the value set in the code.

Could you give me some hints to overcome this bug?

Thank you for all your help again.

With best wishes,
Loc

[Re:11] Bug report

Posted on : June 06, 2017 (Tue) 13:17:49

by 池田敦俊

Dear Mr. DUONG,

I know this topic is rather old, but since no clear answer is written, I will show my solution.
Probably I had the same problem as yours.
Some parts of my .spc file is like this
56.9546721 94.9028466 165.5205792 292.3944519
531.97205921109.00695121118.37976341004.5128776
737.2307654 136.4934815 57.1074870 41.0239120
Spaces are missing in the second line.
This happens when the value has 12 digits or more.

I overcame this problem by changing the 65th line of source/wrtspc.f.
Initially it is
30 write(26+is,'(1x,500f12.7)')(dble(spctrl(k,kp)),kp=1,nk3).
I changed it to
30 write(26+is,'(1x,500f13.7)')(dble(spctrl(k,kp)),kp=1,nk3).
If your .spc file has some value with more than 13 digits, you should change the line to
30 write(26+is,'(1x,500f14.7)')(dble(spctrl(k,kp)),kp=1,nk3).

I hope this post helps.

Sincerely,
Ikeda

[Re:10] Bug report

Posted on : December 08, 2016 (Thu) 10:32:19

by Duc-Long Nguyen

Dear Loc,

Did you by chance enter the file name incorrect, or the file name data/WSe2_bulk_nosoc_radius$rmt may causes the problem?
I would try to change the name to something else which is not containing "$".
With spc, we should use the full file name
ex. : spc WSe2_bulk_nosoc_radiu_up.spc

--
Duc-Long Nguyen
Graduate Student, Molecular Science and Technology Program, TIGP
Institute of Atomic and Molecular Sciences,
Academia Sinica, Taipei 106, Taiwan
Phone +886-979279073

[Re:09] Bug report

Posted on : September 23, 2016 (Fri) 17:13:23

by DUONG, DINH LOC

HI Hitoshi,

Thank you much for your suggestion. I will try and let you know.

With many thanks,
Loc

[Re:08] Bug report

Posted on : September 23, 2016 (Fri) 15:50:48

by Hitoshi GOMI

Hello Loc,

Since I cannot reproduce your result, it is difficult to say something useful for you.
Anyway, spc program is relatively new function in the AkaiKKR package, so that it may not always stable, I guess.

One possibility is to change the input and output format of spc mode.
I usually use irdfmt=1 and iwrfmt=1.
To change the format, please refer to the comments in source/wrtspc.f and source/readk.f and edit source/spmain.f.

Hitoshi

[Re:07] Bug report

Posted on : September 22, 2016 (Thu) 14:30:06

by DUONG, DINH LOC

Hi Hitoshi,

I have tried to compile specx with ifort but it seems the bug is not gone. I still get the same error, not all the time.

Do you have any idea to overcome this issue?

I compile the code with options:
ifort -O2 -mcmodel=medium -qopenmp -o source/gntcs.o -c source/gntcs.f

Thank you for all your help in advance.

With best wishes,
Loc

[Re:06] Bug report

Posted on : September 19, 2016 (Mon) 15:24:22

by DUONG, DINH LOC

Hi Hitoshi,

I am sorry for my late reply. There is a Chuseok holiday in Korea last week.

I see. It may be from my compilling process.
I just copy the out-put data for your reference. There is some line, the data is overlap.

27.5260032 28.5976672
38.4641603 40.2747937
58.0989492 61.5544419
99.1663542 107.1138644
208.7729055 233.2697297
596.6782518 677.7071792
1099.02230671088.5392774
554.0093133 480.7735708
236.3612845 220.4424093
206.9148682 214.4315339
415.5641349 446.9597044
571.3279090 533.1356930
223.8929920 211.9945282
99.1679615 96.2745101
62.6089667 62.6829364
51.4694476 53.9435628
54.4458220 61.3125215
77.1505579 97.4315963
159.9109512 237.3515792

Do you have any suggestion to overcome this issue?

With many thanks,
Loc

[Re:05] Bug report

Posted on : September 14, 2016 (Wed) 00:59:40

by Hitoshi GOMI

Hello Loc,

> Do you mean you can produce the result and it can be read by spc command?

Yes.

Hitoshi

[Re:04] Bug report

Posted on : September 13, 2016 (Tue) 21:06:25

by DUONG, DINH LOC

Hi Hitoshi,

Do you mean you can produce the result and it can be read by spc command?

I am sorry to keep the $ character. That is from my script.
I am checking the convergence with rmt parameter. The $rmt should be 1.

With best wishes,
Loc

[Re:03] Bug report

Posted on : September 13, 2016 (Tue) 20:16:13

by Hitoshi GOMI

Hi Loc,

I could not reproduce your result.
Anyway, it is better to avoid special characters such as $.

Hitoshi

[Re:02] Bug report

Posted on : September 13, 2016 (Tue) 18:51:35

by DUONG, DINH LOC

Dear Hitoshi,

This is my input file.
For scf:
c--------------------WSe2-from Hitoshi------------------------------------
go data/WSe2_bulk_nosoc_radius$rmt
c------------------------------------------------------------
c brvtyp a c/a b/a alpha beta gamma
hex 6.2021, 3.9488, , , , ,
c------------------------------------------------------------
c edelt ewidth reltyp sdftyp magtyp record
0.001 1.5 sra gga91 mag init
c------------------------------------------------------------
c outtyp bzqlty maxitr pmix
update h 200 0.02
c------------------------------------------------------------
c ntyp
2
c------------------------------------------------------------
c type ncmp rmt field mxl anclr conc
W 1 1 0 3 74 100
Se 1 1 0 2 34 100
c------------------------------------------------------------
c natm
6
c------------------------------------------------------------
c atmicx atmtyp
1/3a 2/3b 1/4c W
2/3a 1/3b 3/4c W
1/3a 2/3b 0.6211c Se
2/3a 1/3b 0.3789c Se
2/3a 1/3b 0.1211c Se
1/3a 2/3b 0.8789c Se
c---------------------------------------------------------


For spc: c--------------------WSe2-Hitoshi------------------------------------
spc data/WSe2_bulk_nosoc_radius$rmt
c------------------------------------------------------------
c brvtyp a c/a b/a alpha beta gamma
hex 6.2021, 3.9488, , , , ,
c------------------------------------------------------------
c edelt ewidth reltyp sdftyp magtyp record
0.001 0.5 sra gga91 mag 2nd
c------------------------------------------------------------
c outtyp bzqlty maxitr pmix
quit 8 200 0.02
c------------------------------------------------------------
c ntyp
2
c------------------------------------------------------------
c type ncmp rmt field mxl anclr conc
W 1 1 0 3 74 100
Se 1 1 0 2 34 100
c------------------------------------------------------------
c natm
6
c------------------------------------------------------------
c atmicx atmtyp
1/3a 2/3b 1/4c W
2/3a 1/3b 3/4c W
1/3a 2/3b 0.6211c Se
2/3a 1/3b 0.3789c Se
2/3a 1/3b 0.1211c Se
1/3a 2/3b 0.8789c Se
c------------------------------------------------------------
300
0.0000000 0.0000000 0.0
0.0000000 0.5773503 0.0
0.3333333 0.5773503 0.0
0.0000000 0.0000000 0.0
0.0000000 0.0000000 0.5
0.0000000 0.5773503 0.5
0.3333333 0.5773503 0.5
0.0000000 0.0000000 0.5
c------------------------------------------------------------

Thank you for all in advance.

With best wishes,
Loc

[Re:01] Bug report

Posted on : September 13, 2016 (Tue) 15:06:55

by Hitoshi GOMI

Hello Loc,

Please show me your input to reproduce your problem.

Hitoshi