In order to reduce the size of the event data file, Phoenix provides only a single URL for the first instance of a duplicated event, then provides the URLs for the duplicates in a separate file: these are keyed to the event in the data set. .