Antonio-Cruciani / Link-stream-diameter

Code associated with the paper: M. Calamai, P. Crescenzi, A. Marino "On Computing the Diameter of (Weighted) Link Streams".

Link stream diameter

Code associated with the paper: M. Calamai, P. Crescenzi, A. Marino "On Computing the Diameter of (Weighted) Link Streams".

Usage

All files needed to generate stats are in the folder ./src/stats/.

Dataset

The link at the dataset is here: https://bit.ly/2TYfdaR
Once unzipped, in order to use the facilities we provide below, the folder graphs must be placed in the folder ./src/stats/. Each file in the folder graphs is a link stream, i.e. a list of temporal edges u v t lambda (see the paper). The links specified in each link stream are (and must be) sorted in increasing order of t.

Generate stats

The list of paths of the link streams to be considered must be written in a input file that must be passed in input. We provide the following files, fileA and fileB, which allow to run our experiments on the dataset we used and that can be downloaded using the link below.

fileA: https://bit.ly/3fEVEMi
fileB: https://bit.ly/2MEsbXQ

The file fileA and fileB assume that the folder graphs is placed in the folder ./src/stats/. Each line of the input file, as it can be seen in fileA and fileB, respect this format: <path of the link stream> <[dummy]> <[0]> : where dummy is an optional integer value that represents the last not dummy node in the graph, as we assume that all dummy nodes have index greater or equal than dummy. We use a negative value like -1 if there aren't dummy nodes in the graph. This is to deal with the transformations described in Section 1.2 of the paper. The last value of the line ([0]) is optional and if there is, the graph is undirected.
We also use the simplest form <path of the link stream> if it is a directed link stream and there aren't dummy nodes in the graph.

Computing Lower bounds:

cd ./src/stats/

• EAT: run python3 double_sweep_stats.py EAT fileA
• LDT: run python3 double_sweep_stats.py LDT fileA
• FT: run python3 double_sweep_stats.py FT fileA
• ST: run python3 double_sweep_stats.py ST fileB

Computing the exact diameter for EAT and LDT using our algorithm:

cd ./src/stats/

• EAT: run python3 rub_diameter_stats.py EAT fileA
• LDT: run python3 rub_diameter_stats.py LDT fileA

Computing the exact diameter using the text-book algorithm:

cd ./src/stats/

• EAT: run python3 text_book_diameter_stats.py EAT fileA
• LDT: run python3 text_book_diameter_stats.py LDT fileA
• FT: run python3 text_book_diameter_stats.py FT fileA
• ST: run python3 text_book_diameter_stats.py ST fileB

Computing pivot diameter:

cd ./src/stats/

• EAT: run python3 pivot_ifub_diameter_stats.py EAT NumHubs NumTimes fileA
• LDT: run python3 pivot_ifub_diameter_stats.py LDT NumHubs NumTimes fileA
• FT: run python3 pivot_ifub_diameter_stats.py FT NumHubs NumTimes fileA
• ST: run python3 pivot_ifub_diameter_stats.py ST NumHubs NumTimes fileB