How to run multi-source RCA benchmark

[Nugine]

It seems that the current open-sourced code only uses metrics data.

RCAEval/main.py

Lines 135 to 138 in 6af507a

	# prepare input paths
	data_paths = list(glob.glob(os.path.join(dataset, "**/data.csv"), recursive=True))
	if not data_paths:
	data_paths = list(glob.glob(os.path.join(dataset, "**/simple_metrics.csv"), recursive=True))

Since RCAEval 1.0 mentions multi-source RCA, is there any guide for benchmark?

[phamquiluan]

Thanks for reporting this issue, @Nugine. I'm working on other projects atm. I'll update the multi-source feature maybe next week.