Hey,
while working on step 2 of milestone 2 (https://github.com/mam10eks/ir-lab-sose-2023/tree/main/milestone-02) we encountered a problem trying to render our local run.txt in the webbrowser as a html file.
Upon running the following tira-run command in the command line we were experiencing the error below.
tira-run --output-directory ${PWD}/bm25-output --image python-never-heard-of-qrels --allow-network true --command “diffir --dataset iranthology-python-never-heard-of --web $outputDir/run.txt > $outputDir/run.html”
1
Traceback (most recent call last):
File “/usr/bin/diffir”, line 8, in
sys.exit(main())
File “/usr/lib/python3.8/site-packages/diffir/run.py”, line 71, in main
diff(args.runfiles, config, cli=args.cli, web=args.web)
File “/usr/lib/python3.8/site-packages/diffir/run.py”, line 86, in diff
html = task.web(runs)
File “/usr/lib/python3.8/site-packages/diffir/run.py”, line 564, in web
json_data = self.json(*runs)
File “/usr/lib/python3.8/site-packages/diffir/run.py”, line 412, in json
diff_query_objects = self.create_query_objects(
File “/usr/lib/python3.8/site-packages/diffir/run.py”, line 138, in
create_query_objects
qrels = dataset.qrels_dict()
File “/usr/lib/python3.8/site-packages/ir_datasets/formats/base.py”, line 193,
in qrels_dict
for qrel in qrels_handler.qrels_iter():
File “/usr/lib/python3.8/site-packages/ir_datasets/formats/trec.py”, line 433,
in qrels_iter
yield from self._qrels_internal_iter(self._qrels_dlc)
File “/usr/lib/python3.8/site-packages/ir_datasets/formats/trec.py”, line 449,
in _qrels_internal_iter
raise RuntimeError(f’expected 4 columns, got {len(cols)}')
RuntimeError: expected 4 columns, got 6
It was expecting four columns of input but apparently was given 6. Is this due to our run.txt having 6 different values in each line? What are we supposed to do now?
Our run.txt looks like this (I only copied the first three of the many, many lines)
1 0 2016.sigirconf_conference-2016.224 1 20.548162321552805 BM25
1 0 2013.sigirconf_conference-2013.184 2 20.44424889331691 BM25
1 0 2009.cikm_conference-2009.1 3 20.42082291927 BM25
Can you also please confirm that our next steps would be to evaluate the effectiveness of our relevance judgments and sending you the qrels as well as the effectiveness score.
Best regards
Python never heard of him