SCM

Forum: support

Monitor Forum | Start New Thread Start New Thread
RE: nops_eval: Error in x[seq(round(rb * d[1L]), d[1L]) ... [ Reply ]
By: Achim Zeileis on 2018-12-25 22:06
[forum:46476]
OK, I see. The problematic code is in the internal function trim_nops_scan() that is used to read the scanned (and possibly converted) PNG image and shave off the margins.

Mainly, this code is called inside nops_scan(). However, it is also leveraged in nops_eval if (1) there are inconsistencies between the scanned data (i.e., Daten.txt) and the metadata (register .csv and solutions .rds) and (2) the argument interactive is TRUE (default).

Thus, you should run nops_eval(..., interactive = FALSE). This will report the inconsistencies so that you can fix them manually in the Daten.txt. And then you can re-run nops_eval().

Good luck!

RE: nops_eval: Error in x[seq(round(rb * d[1L]), d[1L]) ... [ Reply ]
By: Jonathan Jupke on 2018-12-24 10:33
[forum:46473]
It is nops_eval(), I just checked again.
Indeed I messed up the scans and the lower third is missing. Since I do not have access to the exams over the Christmas break I just edited the Daten.txt file manually, hoping that evaluation would still work. Is there a workaround or do I just need to scan again?

Cheers,
Jonathan

RE: nops_eval: Error in x[seq(round(rb * d[1L]), d[1L]) ... [ Reply ]
By: Achim Zeileis on 2018-12-23 00:22
[forum:46472]
That's nops_scan, not nops_eval, I guess? And then it's the part of the code responsible for finding the bottom scanner markings. So maybe these cannot be found?

But even then the warning is somewhat strange. Could you share the PNG that nops_scan() tries to read here? Maybe you can blacken the name/id in case this is a real case. Or you can send it via e-mail to me personally.

nops_eval: Error in x[seq(round(rb * d[1L]), d[1L]) ... [ Reply ]
By: Jonathan Jupke on 2018-12-22 15:06
[forum:46471]
Running nops_eval() returns the following error:

Error in x[seq(round(rb * d[1L]), d[1L]), seq(1, round(0.17 * d[2L]))] :
only zeros may be mixed with negative indicies. (My translation from German)

What might the problem might be?

Best Wishes,
Jonathan Jupke

Thanks to:
Vienna University of Economics and Business Powered By FusionForge