Due Date: Mar 25, before midnight (11:59:59PM)

This assignment comprises both written questions and implementation-based lab.


Answer the following questions from the DPV book i) Q4.4 ii) Q4.5, iii) Q4.9 (consider the cases both with and without negative cycles), iv) Q4.12.

Lab7: Dijkstra's Algorithm and Binary Heap Priority Queue

Implement Dijkstra's algorithm described in Fig 4.8. You must implement your own binary heap priority queue for supporting the deletemin and decreasekey operations required in the algorithm. You cannot use any other python module or data structure other than array, list, set or dict. Note that you must use the prev values to peint the actual shortest paths along with the length.

Before you implement the binary heap version of the priority queue, you can test the correctness of the Dijkstra implementation using the basic array based priority queue.

To next implement the binary heap, you can use an array-based implementation of the binary tree (indexed from 0 as usual), there the root is at position $0$ always, and for any position $i$, its parent node is at position $(i-1)//2$, and its children are at positions $2i+1$ (left) and $2i+2$ (right). See Figure 4.16 (and question 4.16) for more details on this array-based implementation of the priority queue. However, you can also implement a non-array based implementation of binary heap if you choose (e.g., via dict, etc., which will be less efficient than the array based implementation).

You should run your code on the following graph: graph1 Each line in the file is a source-target-weight triple separated by spaces. You should treat all nodes and weights as integers. Submit the output for source vertex 0.

Given a source node $s$, your code should print out the length of the shortest path from $s$ to each node $u$ and (one of) the actual shortest paths from $s$ to $u$. So given the graph:

0 1 4
0 2 2
1 2 3
1 3 2
1 4 3
2 1 1
2 2 4
2 3 5
4 3 1

For source node $0$, the output should be as follows:

0 0 [0]
1 3 [0, 2, 1]
2 2 [0, 2]
3 5 [0, 2, 1, 3]
4 6 [0, 2, 1, 4]

Which shows the node, length of shortest path, and actual shortest path per line.

Here is another graph file graph2 and the output using source 30 graph2 output.


Use submitty to submit a PDF and txt and one .py script. The .py file should take the file name and source vertex id from the command line.

The PDF file must contain your solutions to the HW, and the output from the python script can be submitted in the txt file. Show your output for the source vertex 0.

Policy on Academic Honesty

You are free to discuss how to tackle the assignment, but all coding must be your own. Please do not copy or modify code from anyone else, including code on the web. Any students caught violating the academic honesty principle will get an automatic F grade on the course and will be referred to the dean of students for disciplinary action.