sessionize.malloynb

>>>markdown
# Sessionized Data - Map/Reduce

Flight event data contains _dep_time_, _carrier_, _origin_, _destination_ and _tail_num_  (the plane that made the flight).  The query below takes the flight event data and maps it into sessions of _flight_date_, _carrier_, and _tail_num_.  For each session, a nested list of _flight_legs_ by the aircraft on that day.  The flight legs are numbered.
>>>malloy
#(docs) size=large limit=5000
run: duckdb.table('data/flights.parquet') extend {
  where: carrier = 'WN' and dep_time ? @2002-03-03
  measure: flight_count is count()
} -> {
  calculate: session_id is row_number()
  group_by:
    flight_date is dep_time.day
    carrier
    tail_num
    aggregate: 
      flight_count 
      max_delay is max(dep_delay)
      total_distance is distance.sum()
  nest: flight_legs is  {
    order_by: 3
    calculate: flght_leg is row_number() 
    group_by:
      tail_num
      dep_minute is dep_time.minute
      origin
      destination
      dep_delay
      arr_delay
  } 
}
>>>markdown