%sh
DIRECTORY=citibike
FILENAME=201612-citibike-tripdata

echo "Data Set name (remove .zip or .csv)  :" $FILENAME
echo "-----------------------------------------------------------------"
echo ".."
echo "If this paragraph hangs on the SELECT count(*) FROM bike_trips, see the next paragraph for an explanation/resolution...."
echo ".."

cd $DIRECTORY

# run hive
hive <<EOF
DROP TABLE IF EXISTS bike_trips;

CREATE TABLE bike_trips ( 
TripDuration int,
StartTime timestamp,
StopTime timestamp,
StartStationID string,
StartStationName string,
StartStationLatitude string,
StartStationLongitude string,
EndStationID string,
EndStationName string,
EndStationLatitude string,
EndStationLongitude string,
BikeID int,
UserType string,
BirthYear int, 
Gender int
) 
ROW FORMAT delimited 
FIELDS TERMINATED BY ',' ;

LOAD DATA LOCAL INPATH '$FILENAME.nh.csv' into table bike_trips;

create table bike_trips_small as select * from bike_trips limit 50000;

SELECT count(*) FROM bike_trips;

exit;

EOF