Pig script to calculate average distance by airline route

airline.pig

--Load data from view to use
air = LOAD 'default.vw_airline' USING org.apache.hcatalog.pig.HCatLoader();

--Use FOREACH to limit data to origin, dest, distance
--Concatentate origin and destination together, separated by a pipe
--CONCAT appears to only allow two arguments, which is why the function is called twice (to allow 3 arguments)
origindest = FOREACH air generate CONCAT(origin, CONCAT('|' , dest)) as route, distance;

--Group origindest dataset by route
groupedroutes = GROUP origindest BY (route);

--Calculate average distance by route
avg_distance = FOREACH groupedroutes GENERATE group, AVG(origindest.distance);

--Show results in Pig shell
dump avg_distance;

--Write out results to text file, separated by tab (default)
store avg_distance into '/user/hue/avg_distance';