GraphRanker is a C project realized for the Algorithms and Principles of Computer Science (Algoritmi e Principi dell'Informatica) course at Politecnico di Milano in 2021. The focus is on the algorithm's runtime and space optimization.
The goal is to generate a rank of weighted directed graphs. More precisely, the rank has to keep track of the k "best" graphs.
The program reads from the standard input (see following sections for more details):
- two parameters, only once (on the first row of the file)
- d -> number of nodes of every graph
- k -> lenght of the rank
- a sequence of commands among:
- AggiungiGrafo (meaning AddGraph) followed by adjacency matrix
- TopK
- This command indicates that a new graph has to be added to the ones that have to be considered to generate the ranking. It's followed by the adjacency matrix of the graph itself.
- The nodes of the graph are labelled with an integer index between 0 and d-1.
- The weights of the graph have values in the interval [0, 232 - 1]
- Every graph from the beginning of the program to the TopK command has to be considered. Every graph has to be labelled with an integer index corresponding to the number of graphs that have been read before it.
- TopK requires that the k indeces are printed of the graphs with the lower values according to the following metric:
Sum of the shortest paths between node 0 and every other node of the graph - If there are graphs with the same metric value, their priority is assigned according to the reading order (meaning that, the ones who were read first have higher priority).
- The k indeces have to printed in a single line in any order.