Lab 10: Sorting

Objectives:

The main goals for this lab are for you to get more comfortable with sorting algorithms.

Paired Programming rules

This lab is a paired programming assignment. What exactly does that mean? You will be working in pairs on the CS lab computers. Each pair will be working on one computer. One person will be the driver and the other person will be the navigator. Here is the rule: the driver controlls the lab computer, but the driver can only type what the navigator tells them to type. For this to work well, each pair should be constantly talking among themselves. After each problem, you will switch roles, the navigator will become the driver and the driver will become the navigator.

New Partners This week you can switch partners. You will work with your partner this week and next. Your can work with a partner who you have worked with in the past.

Finishing the lab

Before leaving the lab, make sure you fill out the attendance sheet and go through your answers with a TA or instructor. You will not get full credit otherwise.

1. Bubble Sort

In this part of the lab, you will implement Bubble Sort. But first, we will walk through another example together. Slide 21 from lecture 21 has the psuedocode/algorithm for Bubblesort. I’d recommend you pull it up as a reference.

1.1 Example 1

Consider the following array of just 4 numbers: 9,8,7,6. We are going to walk through BubbleSort by hand before we start implementing the algorithm. We will be using a table to track each step of the algorithm and what the array looks like after each step.

Question 1: Fill in just the len and j columns below. len represents the variable in the outerloop and j represents the variable in the innerloop. Do not fill in any cells in the array column yet.

Hint: You might not need every row below.

len	j	array
4	1

Solution

len	j	array
4	1
	2
	3
3	1
	2
2	1

Question 2: Now fill in the values in the array column. Each value should show what the array now looks like after len and j are specific values

len	j	array
4	1	8976

Solution

len	j	array
4	1	8 9 7 6
	2	8 7 9 6
	3	8 7 6 9
3	1	7 8 6 9
	2	7 6 8 9
2	1	6 7 8 9

Question 3: How many swaps do you have to make? How many pairs of elements did we consider?

Solution

we made 6 swaps. We have to compare every pair. We compare: 9 to 8, 7, 6 8 to 7, 6 7 to 6

1.2 Example 2

Now consider the following array of 5 numbers: 5,10,6,2,7. We are going to walk through BubbleSort by hand before we start implementing the algorithm. We will be using a table to track each step of the algorithm and what the array looks like after each step.

Question 4: Fill in just the len and j columns below. len represents the variable in the outerloop and j represents the variable in the innerloop. Do not fill in any cells in the array column yet.

Hint: You might not need every row below or you might need to add rows.

len	j	array

Solution

len	j	array
5	1
	2
	3
	4
4	1
	2
	3
3	1
	2
2	1

Question 5: Now fill in the values in the array column. Each value should show what the array now looks like after len and j are specific values

len	j	array

Solution

len	j	array
5	1	5 10 6 2 7
	2	5 6 10 2 7
	3	5 6 2 10 7
	4	5 6 2 7 10
4	1	5 6 2 7 10
	2	5 2 6 7 10
	3	5 2 6 7 10
3	1	2 5 6 7 10
	2	2 5 6 7 10
2	1	2 5 6 7 10

Question 6: How many swaps do you have to make? How many pairs of elements did we consider?

Solution

we made just 5 swaps: (5,2), (5,3), (5,4), (4,2), (3,1) We have to compare every pair. We compared: 10 to 5, 6, 2, 7 5 to 6 6 to 2, 7 2 to 5 5 to 6 2 to 5 thus we made 10 comparisons

Question 7: When we had an array of size 4, we made 6 comparisons, when we had an array of size 5, we made 10 comparisons.

Imagine we have an array of size n. How many comparisons will we have to make?

Solution

$$\dfrac{n * (n-1)}{2}$$ This is because we compare each element to every other element. thats where we get the $$n(n-1)$$. We can divide that number by 2 since we just care about comparing each pair. For example, we care about comparing A and B, so if we consider A-B, we dont need to check B-A. Therefore, the total number of comparisons we need to make is $$\dfrac{n^{2} - n}{2}$$

Question 8: In what scenario would we make the fewest number of swaps? In what scenario would we make the most number of swaps?

Solution

If the list was sorted (in ascending order) then we dont need to make any swaps. If the list was sorted (in descending order) then we need to swap whenever we compare two items in the list.

1.3 Implementation time

We have provided starter code in Bubble.java for you to implement. You can access the code here.

There are two methods for you to implement, swap() and bubbleSort().

Solution

We have provided the solution in BubbleSolution.java. You can access the code here.

2. Selection Sort

In this part of the lab, you will implement Selection Sort. But first, we will walk through another example together. Slide 49 from lecture 21 has the algorithm for selection sort. I’d recommend you pull it up as a reference.

2.1 Example 1

Consider the following array of just 4 numbers: 9,8,7,6. We are going to walk through SelectionSort by hand before we start implementing the algorithm. We will be using a table to track each step of the algorithm and what the array looks like after each step.

Question 9: Fill in the startIdx, minIdx, minVal, and array columns below. startIdx represents the variable in the outerloop, minIdx represents the index of the smallest value in the remaining parts of the list, minVal represents the value at minIdx, and array represents the list after the minimum value has been moved to the current startIdx (if needed).

Hint: You might not need every row below or you might need to add a row.

startIdx	minIdx	minVal	array
0
1
2
3

Solution

startIdx	minIdx	minVal	array
0	3	6	6 8 7 9
1	2	7	6 7 8 9
2	2	8	6 7 8 9

Question 10: How many swaps do you have to make? How many pairs of elements did we consider?

Solution

Just 2 swaps, 6 with 9 and then 7 with 8. Comparisons, is a bit tricky. We are making comparisons when we find the minIndex. When startIndex is 0, we have to compare 4 elements (9,8,7,6), when startIndex is 1, we have to compare 3 elements (7,8,9). When startIndex is 2, we have to compare 2 elements (8,9). When startIndex is 3, we dont need to compare any elements because the reminaing list of unsorted numbers has just one number. In total this is 4 + 3 + 2 = 9

2.1 Example 2

Consider the following array of 5 numbers: 5,10,6,2,7. We are going to walk through SelectionSort by hand before we start implementing the algorithm. We will be using a table to track each step of the algorithm and what the array looks like after each step.

Question 11: Fill in the startIdx, minIdx, minVal, and array columns below. startIdx represents the variable in the outerloop, minIdx represents the index of the smallest value in the remaining parts of the list, minVal represents the value at minIdx, and array represents the list after the minimum value has been moved to the current startIdx (if needed).

Hint: You might not need every row below or you might need to add a row.

startIdx	minIdx	minVal	array
0
1
2
3

Solution

startIdx	minIdx	minVal	array: 5, 10, 6, 2, 7
0	3	2	2 10 6 5 7
1	3	5	2 5 6 10 7
2	2	6	2 5 6 10 7
3	4	7	2 5 6 7 10

Question 12: How many swaps do you have to make? How many pairs of elements did we consider?

Solution

Just 3 swaps, 2 with 5, 5 with 10, 7 with 10 Comparisons, is a bit tricky. We are making comparisons when we find the minIndex. When startIndex is 0, we have to compare 5 elements, when startIndex is 1, we have to compare 4 elements. When startIndex is 2, we have to compare 3 elements (8,9). When startIndex is 3, we compare 2 elements. When startIndex is 4, we dont need to compare any elements because the reminaing list of unsorted numbers has just one number. In total this is 5 + 4 + 3 + 2 = 14

SelectionSort also is an quadratic algorithm, meaning that as the size of the list increases, the amount of steps the algorithm will take in the worst case scenario increases a squared amount of times. We will discuss this in more detail on Thursday or next week.

The idea of how long an algorithm will take (or how many steps it must perform) in the worst case scenario is an important concept in Computer Science. It will be covered in detail in Data Structures and Discrete Math!

2.3 Implementation time

We have provided starter code in Selection.java for you to implement. You can access the code here.

There are two methods for you to implement, swap() and selectionSort().

Solution

We have provided the solution in SelectionSolution.java. You can access the code here.

Wrap up

In today’s lab we covered two sorting algorithms. This will help you for your last homework.

Signing out

Before leaving, make sure your TA/instructor have signed you out of the lab. If you finish the lab early, you are free to go.