octave_base_value::array_value(): wrong type argument `complex matrix' error

[Vahid Askarpour]

Dear USPEX Users,

Since there was no response to my "structure has no member 'molecules' " post, I installed an earlier version of octave (3.0.5) to resolve the problem. After two generations, when I remove the still_reading file and restart octave, I get the following in my output:

error: octave_base_value::array_value(): wrong type argument `complex matrix'
error: evaluating for command near line 18, column 1
error: evaluating if command near line 8, column 1
error: called from `writeCalcFilesMOL_SIESTA' in file `/home/vahid/scratch/USPEX-9.1.0/Code/FunctionFolder/USPEX/writeCalcFilesMOL_SIESTA.m'
error: evaluating if command near line 349, column 1
error: evaluating if command near line 299, column 1
error: evaluating for command near line 286, column 1
error: evaluating if command near line 155, column 1
error: evaluating while command near line 154, column 1
error: evaluating if command near line 127, column 1
error: evaluating while command near line 121, column 1
error: evaluating while command near line 10, column 1

In the log file I simply get ans=0

This error seems to arise when using molecular regime with siesta. It does not happen with VASP.

I appreciate any hints or suggestions.

Vahid

[Andriy Lyakhov]

Dear Vahid,

it looks like some octave related bug. If you still have the calculation files or will encounter it in the future, can you pelase send me the files Current_POP.mat, Current_ORG.mat and Zmatrix.data from the CalcFolder?

Sincerely,

  Andriy

[Vahid Askarpour]

Dear Andriy,

Attached please find the files you requested. The Zmatrix.data are labeled from 1 to 10 as there are ten parallel calculations.

Thank you for looking into this and for your response on the Test T13 thread.

Vahid

[Andriy Lyakhov]

Hello,

your files look fine, no abnormal or complex values of the variables used in that .m file. One bug was reported on octave forums that causes message like that a few months ago and it might be related to that bug.

Please remove the ; at the end of the lines number 19-23 in the file writeCalcFilesMOL_SIESTA.m and try to continue the USPEX calculation. When it stops with the same error message, I will need the last lines from your log file, maybe I will understand what is going on.

Sincerely,

  Andriy

[Vahid Askarpour]

Hi Andriy,

The end of the log file is as follows:

ans = 0

nowMOL =  1

nowMOL =  1

format =

   0   0   0

   1   0   0

ZZZ =

   2.7397 + 0.0000i   6.9123 + 0.0000i   6.2316 + 0.0000i

   1.6704 - 0.0000i   1.2659 + 0.0000i  -2.4669 + 0.0000i

infoFFile =

 Columns 1 through 3:

   1.00000 + 0.00000i   0.00000 + 0.00000i   0.00000 + 0.00000i

   2.00000 + 0.00000i   1.00000 + 0.00000i   0.00000 + 0.00000i

 Columns 4 through 6:

   0.00000 + 0.00000i   2.73967 + 0.00000i   6.91231 + 0.00000i

   0.00000 + 0.00000i   1.67044 - 0.00000i   1.26594 + 0.00000i

 Columns 7 through 9:

   6.23165 + 0.00000i   1.00000 + 0.00000i   1.00000 + 0.00000i

  -2.46690 + 0.00000i   1.00000 + 0.00000i   1.00000 + 0.00000i

 Column 10:

   1.00000 + 0.00000i

   1.00000 + 0.00000i

Sincerely,

Vahid