Extracting random structures from AIMD's vasprun.xml

[stemdkk]

Dear Dr. Tadano and users,
I am using displac.py to extract random displacement structures from AIMD. When I am simultaneously performing VASP6 on-the-fly machine learning during AIMD, the generated vasprun.xml produces the following error:The
range specified by --every is larger than the loaded MD data. I checked and found that out of a total of 4000 steps, 126 steps are DFT-calculated structures, while the rest are predicted structures. When I change the range to 1:120:20, I can get results; however, using 1:130:20 gives an error. I suspect it might be an issue with the AIMD format, but I don't know how to check or modify it

[ttadano]

Hi,

Thank you for the report. Could you tell me how many snapshots were extracted when you run extract.py without --every option?

[stemdkk]

AIMD has a total of 4000 steps,--every=1001:3000:25, a total of 80 snapshots here. Thank you.

[ttadano]

Thank you. While I am not sure, the error occurred probably because the format of vasprun.xml differs when you use the on-the-fly machine-learning option.

[stemdkk]

Thank you for your reply. Because I have a lot of systems to calculate, how can I speed up my AIMD calculations on the basis of accuracy?