Issue with lhis limit when calling hpts for large number of history points

[Chinthaka Jacob]

Dear Nek Community,

I am encountering an issue that at first seemed straightforward but has proven persistent despite some troubleshooting. I am working with a genbox-generated domain (later modified via the usrdat2 subroutine, but still tensor-product structured), and I need to average the solution fields along the z-direction in post-processing. To achieve high spatial resolution, I am using the hpts tool to interpolate the mesh and have defined a large number of history points, on the order of 10⁷.

However, the job crashes with the following message, even after increasing lhis in the SIZE file:

dump history points 

reading history points 

ABORT: Increase lhis in SIZE!

Despite setting lhis to a large value, the problem persists. Has anyone encountered this before, or can suggest a workaround? I have attached the relevant logfile for reference.

Thanks in advance for your time and insights.

Chinthaka.

[Philipp Schlatter]

I would go to the code and see where the comment is written out. Then check.what conditions in the if statement is causing the problem. This would.certainly help understanding the issue.

Btw, have you made sure to clean and recompile the code? Another comment, if this is postprocessing, you may be interested in PySEMTools to be able to do these things in Python afterwards.
Best regards, Philipp

--
Sent from my Android device with K-9 Mail. Please excuse my brevity.

[Chinthaka Jacob]

Many thanks, Philipp, for the helpful information! I revisited the relevant part of the code and traced the specific condition triggering the error. Also, your comment regarding PySEMTools is particularly valuable — I wasn’t aware of this postprocessing capability...

Chinthaka.