Study the impact of O() on a problem solution and learn to design and tweak an experiment (in java):

1 - Use the 3 different algorithms for Maximum Contiguous Subsequence Sum, you can copy the code from http://www.crazyforcode.com/maximum-sum-in-contiguous-subarray/ - Give credit. You will need to tweak the code to make it work. You can also simplify the code...

2 - Use StopWatch class from the book for timing

Book's StopWatch example:

public class Stopwatch {

private final long start;

public Stopwatch() {

start = System.currentTimeMillis();

}

public double elapsedTime() {

long now = System.currentTimeMillis();

return (now - start) / 1000.0;

}

}

3 - Build a tester that: * Generates datasets randomly * Run all 3 algorithm with each generated dataset * You will need multiple test-runs to establish N such that value of T(N) are neither too short (Resolution is mili-second) or too long (Longer than ~ 1 day) * Initially loop multiplying the dataset size by 2 (Initial multiplier) every time

4 - Follow the analysis of experimental data from the book page 176. Plot the data using excel

books:

5 - From this estimate a good starting N and increment such that your table will have at least 4 lines with all 3 results

6 - Run the test and repeat 10 times to obtain an average (you can test with other work-load present or computer idle)

7 - Build a table that shows N vs. average / minimum / maximum T(N) for all 3 algorithms

8 - Turn a professional looking report: a. describes the experiment b. include the code c. include a sample run d. explain the final set of numbers you used (starting size and multiplier and time cut-off) e. explain the findings f. estimate the constant factors for all 3 algorithms assuming aN-cube, bN-square, and cN.

9 - code (.java) and the report is needed