I ran into this issue recently where participation_group was repeated in hh, person, day, trip and vehicle tables.
It really confused hts_summary -- it ended up choosing the lowest-grain table (trip) for summaries, so I ended up with wackadoo counts of each participation group level.
We should make sure that the tables fed to hts_data follow database design principles: unless a key column (IDs), no column should be repeated across tables.
I ran into this issue recently where participation_group was repeated in hh, person, day, trip and vehicle tables.
It really confused hts_summary -- it ended up choosing the lowest-grain table (trip) for summaries, so I ended up with wackadoo counts of each participation group level.
We should make sure that the tables fed to hts_data follow database design principles: unless a key column (IDs), no column should be repeated across tables.