Good Day Xeekers!
Thank you for participating in Off the Beaten Path. This is a new challenge style for us, and we are eager to see how you approach it.
In recent testing, we found a bug in the Lyra Graphtool that shows up only in some fringe graph-building workflows. You are unlikely to see this bug. We’ve updated the Lyra Graphtool on Train data for the challenge for good measure.
Let us know if you have questions about the Lyra Graphtool. Our crew is interested in making it better and using it as a basis to build more exciting challenges.
Thank you for sharing such a tool, it will be very helpful for the competition.
I noticed a bug in the current code, the load function didn’t seem to take into account the graph_file parameter, so I looked into it and found that it was because it is not updating the
args_list variable (it is updating the
args_dict variable instead.
You are missing this line before the return statement :
args_list = json_to_list(args_dict)
I confirm the issue.
In the meantime, I’m loading the graph afterwards.
pargs = lgtool.ProcessArgs.load(arguments_file="args_leaderboard")