Many relative to the total set, no. But enough to know writing it down isn’t going to cut it. Most datasets I’ve seen are xls, hdfs or csvs nested many directories deep. Not exactly for those who can’t manage an email filter. Without getting into the processing of it either