The current data from the timetable API has Sydney Train shapes that are invalid due to the point_sequence values being wrong.
For an example - see the attached image of shape_id ‘CTY_S1d’ with point sequence labels.
I can’t see any workaround for this as the raw file has the records in that order - does this happen often?