Duplicate Raw IRIS Data

[ntjohnson1]

The raw data appears to be duplicated for some episodes in the IRIS dataset.
For instance this and this have the same uuid but different directories.

These seem to be isolated to IRIS and the duplicated versions have the H_M_S format instead of the H:M:S format seen elsewhere.

[ntjohnson1]

A fairly simple query on the filenames find there are about 50 duplicates all in IRIS all on the same day (some samples in REAL use the H_M_S format but don't have duplicates. I didn't verify ALL the uuids

dedup_sample = [sample for sample in filepaths if sample.replace(":","_") == sample or sample.replace(":","_") not in sample_set]