log4j error when using sync_reasoner_pellet

[isukura]

Hello,

my code structure is something like this:

for owl_file in [a.owl, b.owl]:
    ... (some ontology work)
    sync_reasoner_pellet(ontology)
    ...

When iterating in this order, [a.owl, b.owl], result is:
    1. code passes through, result is as expected for a.owl
    2. code passes through, but result for b.owl is totally weird

When iterating in different order, [b.owl, a.owl], result is:
    1. code passes through, result is as expected for b.owl
    2. code doesn't pass through for a.owl. Java error: error.txt

Therefore, I have two questions.
1. Why does code behave differently depending on the order? Are maybe some global variables defined which I'm not aware of, that could affect the results?
2. Do you maybe know what could be an issue that causes the Java error? I scrolled through old posts, but unfortunately did not figure it out. I read that it might be related to insufficient Java memory, but due to the 'Could not reserve enough space for 2048000KB object heap' error, I already had to reduce the memory variable in reasoning.py, so I can't increase it.

Thank you in advance!

[Jiba]

Hello,

Another possible explanation is that, when making the reasoning on the first ontology, it somehow modify the second one and thus modify the reasoning.

It which ontology did you asserted the inferred facts? Possibly you can try to put the call to sync_reasoner_pellet() inside a "with onto:..." block. You may create a new, empty, ontology for storing the inferred facts. Then try to make the reasoning on the second ontology. You may also try to destroy the ontology with the inferred facts, in order to see whether it changes something or not.

Do you have a working example that I can use to reproduce the problem?

Jiba