hadoop-accidents

The goal of this project is to infer qualitative data regarding the car accidents in New York City.

TODO

• charts

Input file

The input dataset can be downloaded from here.

How to build

To automatically set env and download sample data

Execute init.sh, this will download csv file and will create a run.sh script explained in the next section

To compile and fetch dependencies

gradle build

How to run

Automatic way

Execute run.sh passing as argument runAll to run all queries or you can select only some of them (passing them as separated arguments) from the following list:

• ContributingFactors
• LethalPerWeek
• WeekBorough

for example ./run.sh LethalPerWeek WeekBorough.

Manual way

# To run the queries on the cluster
hadoop jar build/libs/hadoop-accidents-1.0-SNAPSHOT.jar mw.hadoop.queries.ContributingFactors file:///path/to/NYPD_Motor_Vehicle_Collisions.csv file:///path/to/output1
hadoop jar build/libs/hadoop-accidents-1.0-SNAPSHOT.jar mw.hadoop.queries.LethalPerWeek file:///path/to/NYPD_Motor_Vehicle_Collisions.csv file:///path/to/output2
hadoop jar build/libs/hadoop-accidents-1.0-SNAPSHOT.jar mw.hadoop.queries.WeekBorough file:///path/to/NYPD_Motor_Vehicle_Collisions.csv file:///path/to/output3