1. Problem statement
A sample input file is given below. Each line corresponds to a point-of-interest (POI), which contains a keyword, coordinate values x and y (separated by white space).
park 3 5
lake 2 3
mall 1 4
park 2 4
lake 9 8
mall 2 7
We measure the distance between two points p1=(x1,y1) and p2=(x2,y2) by:
________________
dist(p1, p2) = (x1 – x2)2 + (y1 – y2)2
Each keyword k is associated with a group G(k) of points.
[Example] The group of “park” contains two points: (3,5) and (2,4).
There are 2 questions in this programming assignment. You should write a MapReduce program to solve each of them.
Question Q1: Find the centroid (i.e., the mean position of points) of each group.
[Example]
Input: the sample input above
Output:
lake 5.5 5.5
mall 1.5 5.5
park 2.5 4.5
Question Q2: Find the diameter (i.e., the maximum distance between any two points inside a group) of each group.
[Example]
Input: the sample input above
Output:
lake 8.602
mall 3.162
park 1.414
2. Requirements
• Though MapReduce support multiple languages, in this assignment, you should use Java (Java 8) for implementation.
• You submission should be organized as follows
— Q1.java // source file for question 1
— Q1.jar // jar file for question 1, compiled and archived from Q1.java
— Q2.java // source file for question 2
— Q2.jar // jar file for question 2, compiled and archived from Q2.java
• Make sure that you can compile your source file and run with the latest Hadoop version’s (i.e., Hadoop 3.2.1) pseudo-distributed mode.
• Your jar file should be directly runnable on Linux platform with the following call:
bin/hadoop jar Q1.jar Q1
• Your output result should preserve double precision.
• You should only use one MapReduce round to solve each sub-question.
• [Hint] You may use the Ubuntu image we provided for this assignment.
• Google drive:
https://drive.google.com/file/d/1lMqmTAj2sC2gVqkVWW-MDUR24vv-a3Si/view?usp=sharing