[ERROR][NetCDF] Not a directory

[laurent.pedesseau]

Dear Yambo masters,

I have dowloaded the yambo binary namely yambo-3.2.3-rev11_gnu_core2.tar.gz. I have successfully done the tutorial of LiF with it. Then, I tried to apply it in the same way to my study interest.

Actually, during the last step, I have the following error message:
_______________________________________________________
<---> [05.01] Main loop
<---> BSK | | [000%] --(E) --(X)

[ERROR] STOP signal received while in :[05.01] Main loop
[ERROR][NetCDF] Not a directory
_______________________________________________________

Please find in attached, all the directories and files in tar.gz that you asked for helping me.

Regards,

Laurent PEDESSEAU

[myrta gruning]

Dear Laurent

Looking at your files I cannot spot anything obvious that would give rise to the error you reported. Note that netcdf is useful but also quite nasty, sometimes I got netcdf errors in my clauclations that disappear just deleting the offending file repeating the calculations . Anyway I downloaded your files and will see if I can reproduce the error or understand its origin.

Beside your problem I noticed that in the initialization step you put the MaxGvecs key that in practice reduces (in your case quite drastically) the cutoff with respect the DFT calculations. This will of course speed up the calculations and diminish the memory needed by the calculation and for that reason is used in the tutorial examples that are meant to illustrate how the program works and to get qualitative results.
However it will also affect the final results. So for production run it is better not to use this key (so you will use the default) if not really needed (e.g. memory problems) and in any case its effect on the final results has to be estimated.

Best,
Myrta

[myrta gruning]

Dear Laurent,

I understood what can be your problem
I reproduced the error message by using

>yambo -F 06_BSE -J 06_BSE

Did you use this command line?

now: 06_BSE is the file with your input. With -J you want to create a directory 06_BSE where you put the database of the calculation. Before creating it Yambo checks if is exists already, and in this case it finds out that it exists but it is not a directory.

Note that in the tutorial the input are in a separate directory. So either you move/rename the input or change the suffix you pass with -J

Best,
m

[laurent.pedesseau]

Dear Myrta,

I have actually changed the name of the suffix that I was passing with -J.

And, It is working :


__ __ _______ __ __ _______ _______
| | | || _ || |_| || _ || |
| |_| || |_| || || |_| || _ |
| || || || || | | |
|_ _|| || || _ | | |_| |
| | | _ || ||_|| || |_| || |
|___| |__| |__||_| |_||_______||_______|

<---> [01] Files & I/O Directories
<---> [02] CORE Variables Setup
<---> [02.01] Unit cells
<---> [02.02] Symmetries
<---> [02.03] RL shells
<---> [02.04] K-grid lattice
<---> [02.05] Energies [ev] & Occupations
<---> [03] Transferred momenta grid
<---> [04] External QP corrections (K)
<---> [04.01] QP Corrections extrapolation
<---> [05] Bethe-Salpeter Kernel
<---> [BSE] Kernel dimension : 864
<---> [05.01] Screneed interaction header I/O
<---> Filling Estimation |####################| [100%] --(E) --(X)
<---> [P 01] Kernel filling [o/o] 100.0000
<---> [FFT-BSK] Mesh size: 12 12 18
<---> [WF-BSK loader] Wfs (re)loading |####################| [100%] --(E) --(X)
<---> [05.01] Main loop
<---> BSK |####################| [100%] --(E) --(X)
<---> [06] BSE solver(s)
<---> [06.01] Haydock solver
<---> [Haydock] Iteration 11 Accuracy : 0.01783| -0.02000
<---> [07] Game Over & Game summary


Thank you very much for your time and also for your reactivity.

All the best,

Laurent