231-algorithms

Question 1

Goals of algorithmic design

Answer 1

-time complexity- run as quickly as possible
- space complexity - take up the least amount of memory
- situational. if you have a lot of processing power, time is less important and you would focus on less data wastage (space), if you need a lot of data to be processed quickly, time is important

Question 2

Algorithm

Answer 2

Sequence of steps followed in order to solve a problem/perform particular task

Question 3

Big O Notation

Answer 3

Expresses the complexity (efficiency) of an algorithm in terms of how well it scales as dataset grows

takes in consideration best, average and worst case scenarios

approximation, allows you to predict the memory req/execution time it takes for an algorithm to finish given the data input

Question 4

Big O Notation Rules

Answer 4

Remove all terms except the one with the largest factor or exponent
Remove any constants
e.g O (n^2 + n + 1000) simplifies to O (n^2)

Question 5

Time Complexity

Answer 5

Number of cycles of processing time
running time required in memory depending on amount of data input
Time taken to complete the algorithm/solve the problem
How well an algorithm scales in the processing time taken/number of cycles as the problem dataset grows

Question 6

Space Complexity

Answer 6

running space required in memory depending on scale data input
Amount of storage algorithm takes

Question 7

Reducing Space Complexity

Answer 7

Perform all changes on the original piece of data (copies take space)

Question 8

Reducing Time Complexity

Answer 8

Reduce the number of embedded loops and number of items you have to complete the operations on

Question 9

Best case time complexities

Answer 9

Linear Search - O(1)
Binary Search array and tree - O(1)
Hashing - O (1)
Breadth/Depth first - O(1)
Bubble sort - O(n)
Insertion sort - O(n)
Merge sort - O(n log n)
Quick sort - O(n log n)

Question 10

Average case time complexity

Answer 10

Linear search - O(n)
Binary - O(log n)
Hashing - O(1)
Breadth/Depth - O(V+E)
Bubble sort - O(n^2)
Insertion sort - O(n^2)
Merge sort - O(n log n)
Quick sort - O(n log n)

Question 11

Worst case time complexity

Answer 11

Linear search - O(n)
Binary search array - O(log n)
Binary search tree - O(n)
Hashing - O(n)
Breadth/depth first - O(V^2)
Bubble sort - O(n^2)
Insertion sort - O(n^2)
Merge sort - O(n log n)
Quick sort - O(n^2)

Question 12

Worst Space time complexity

Answer 12

Linear search - O(1)
Binary search array- O(1)
Binary search tree- O(n)
Hashing- O(n)

Bubble sort - O(1)
Insertion sort - O(1)
Merge sort- O(n)
Quick sort- O(n) or log n (average)

Question 13

O (1) Constant Big O Notation

Answer 13

Algorithm will always take the same time/space
Independent from the input size
no loops or iterations

Question 14

O (n) Linear Big O Notation

Answer 14

Time/Space increases in direct proportion to input size
reduced efficiency as dataset grows
for loop/while loop

Question 15

O (n^2) Polynomial Big O Notation

Answer 15

how well the algorithm scales in time taken/memory is directly proportional to square the power of n of the input size

efficiency is significantly reduced as dataset grows.
deeper nested iterations result in higher power

Question 16

O ( log n) Logarithmic Big O Notation

Answer 16

amount of time/memory increases at a smaller rate as the size of input increases
Discards irrelevant data input e.g divide and conquer/halving the data set

Question 17

O (n log n) Linearithmic

Answer 17

divides a data set but can be solved using concurrency or divided lists
has linear and logarithmic growth rate factor to it. e.g
merge sort splits arrays logarithmically, and linear sots when merging them

Question 18

O (2^n) exponential big o

Answer 18

how the algorithm scales is exponential
With each data input, it doubles

recursive

Question 19

O (n!) factorial

Answer 19

multiplies an ever increasing amount for every extra item in the dataset, considers such permutation

Algorithm grows rapidly with each addition to input size. Factorial of input.

Question 20

searching algorithms

Answer 20

to find a specific element within a data structure, performed on a set of data
each algorithm suited to a particular data structure

Question 21

linear search

Answer 21

• starts at beginning of dataset
• traverses through every item one at a time in sequence
• until it finds the item its searching for
  else if all items have been inspected, returns false
• unsorted or sorted
• easier to program/implement (linked list/array)
• not efficient for large lists, ideal for smaller datasets
• adding new item to list doesnt matter

Question 22

linear search pseudocode

Answer 22

.
Function linearSearch(list, item)
index= -1
i=0
found = False
while i < length(list) and found = False
if list[i]= item then
index= i
found= True
endif
i=i+1
endwhile
return index
endfunction

Question 23

binary search

Answer 23

• starts at midpoint
• uses divide and conquer to search for an item
• halves search space with each iteration logical comparison, repeats until midpoint = item. uses 3 pointers: mid, low, high.
• array/list must be sorted
• implemented w/ array or binary tree
• harder to program,
  more complex algorithm, so linear search on small datasets outperform it
  -very efficient for large data sets, needs sorted
• additional time to sort data, and if adding new item

Question 24

binary search iterative pseudocode

Answer 24

binary search code (iterative)
function binarySearch(array,item)
low=0
high=array.length-1
while(low<=high){
mid=(low+high)/2
if (array[mid]>item)
high=mid-1
//discard the high half of the array
else if (array[mid] <item)
low=mid+1
else
return mid
//return index of mid
end if
end while
return -1
//returns -1 if value isn’t found
end function

Question 25

binary search recursive pseudocode

Answer 25

function binarySearch(array,item,low, high) if (highitem) return binarySearch(array,item,low,mid-1) //call BinarySearch on low half of array else if(array[mid]

Question 26

sorting algorithm

Answer 26

designed to take in a number of elements and arrange them in logical order within data structure

Question 27

bubble sort

Answer 27

- loops through every item of array - each item compared to its next, if out of order then swap - process repeated for each adjacent pair - flag variable stores if any swaps in pass when there are no swaps, algorithm ends - each pass, largest number bubbles to the end, no need to check that - ez to implement simple, inefficient for small datasets

Question 28

bubble sort pseudocode

Answer 28

function bubbleSort (array) //pass in an unsorted arrat as parameter n=array.length //get the length of the array swap=True //set an initial flag to be true while (swap=true) //loop as long as swap remains true swap=False //set swap to false to assume no swaps this pass for(i=0 to n-1) //loop from 0 to number of items if (array[i])>array[i+1]) //compare the index to it.s neighbour swap=true //if a swap happens set swap to true temp=array[i] //these lines swap the two values around by array[i]=array[i+1] //storing one in a temporary variable array[i+1]=temp end if next i n=n-1 //reduce the number of items as we know that the one on the end is defintely the biggest end while return array //end by returning the sorted array end function . A = Array of data for i = 0 to A.length - 1: noSwap = True for j = 0 to A.length - 1: if A[j] > A[j+1]: swap A[j] and A[j+1] noSwap = False if noSwap: break return A

Question 29

insertion sort

Answer 29

- by comparing every other in a linear way if value is greater than index, we shift that by 1 and continues until correct insertion position found. - starts at the second index, first value sorted. unsorted and sorted side. - iterate through unsorted. For each element, compare it with the sorted side elements. Shift sorted elements to right that are larger. Insert in correct position -until every item has been inserted -similar to bubble sort but performs better if dataset already mostly sorted, less comparisons

Question 30

insertion sort pseudocode

Answer 30

insertion sort code function insertionSort(array) //aaray passed in as parameter n=array.length //get length of array for i=1 to n-1 ..loop from item 1(not 0!) to length -1 currentValue=array[i] //gets value of current array indexx position=i //sets a variable to keep track of position to loop back while position>0 and array[position-1]> currentValue //sets up the loop backwards array[position[=array[position-1] //moves values to the right if bigger position=position-1 //move left one end while array[position]=currentValue //adds the item in at the correct insertion point next i return array //return the sorted arrat end function . A = Array of data for i = 1 to A.length - 1: elem = A[i] j = i -1 while j > 0 and A[j] > elem: A[j+1] = A[j] j=j- 1 A[j+1] = elem

Question 31

merge sort

Answer 31

- uses divide and conquer and recursion - array halves recursively until individual element - adjacent sub lists compared, merged - inspects first item of both sub lists and smallest is inserted into new list, until both lists are empty - more efficient, can work on multiple lists at the same time, large data sets, - hard to code, high space complexity as subsets need to be held in memory at once - ideal for parallel processing environments

Question 32

merge sort pseudocode section (splitting)

Answer 32

merge sort code function mergeSort(mergeList) //array is passed in as parameter if (mergeList.length<2) //checks the length isn't one return array //if it is return the array end if //this is the terminator for the recursion mid=mergeList.length DIV 2 //gets the midpoint using an integer division leftHalf=mergeList.splice(0,mid) //gets the left half of the array rightHalf=mergeList.splice(mid,mergeList.length) //gets the right half return merge(mergeSorted(leftHalf),mergeSort9rightHalf) //call the merge function on the mergeSort of the left and right halves.. This causes the recursion to trigger end function

Question 33

merge sort pseudocode (merging)

Answer 33

function merge(leftHalf,rightHalf) //this function merges two arrays result=[]; //create an empty array for the result i=0; j=0; k=0; //declare counter variables wile i

Question 34

quick sort

Answer 34

- divide and conquer algorithm - chooses pivot point, elements partitions into sublists according to whether less/greater than pivot - each side recursively sorted, pivot chosen. until each item is individual - uses concurrency to work on each partition at the same time -depends on data being sorted and pivot value - parallel processing environments, real time situations

Question 35

pathfinding algorithms

Answer 35

find optimal/satisfactory paths between vertices and nodes on graphs calculates shortest path

Question 36

Dijsktra shortest path algorithm

Answer 36

- finds shortest path between one node and all other on a weighted graph usage: e/g shortest path for packet routing, GPS navigation, telephone networking - implemented using a priority queue, smallest distances stored at front. breadth first - doesnt work for negative value edges

Question 37

dijkstra shortest path algorithm steps

Answer 37

- make list of visited vertices, unvisited vertices - start vertex, add distances of neighboring nodes to priority queue - visit unvisited vertex with smallest known distance from start - if working distance is less than known distance, update table, mark as visited - ignore routes that are not the shortest - repeat until all visited -backtrack from final vertex

Question 38

A* algorithm (pathfinding algorithm)

Answer 38

Algorithm that finds the shortest path between two nodes uses a heuristic function (value) to optimise performance, approximate cost from node x to final node using extra information - once target node is visited, algorithm stops, not every vertex is considered - It keeps a sorted priority queue of alternate path segments along the way -when deciding on node to visit, looks at heuristic cost + actual cost of unvisited node compared to working total - effectiveness depends on accuracy of heuristics used - quicker, reduce the time taken to find a path - travel routing, video games NPCS, packet routing, financial modelling. pathfinding and graph traversal