Ideas for graphs/analysis

[simonedegiacomi]

What type of analysis should we do? what type of graphs should we create?

Possible ideas for graphs:

Latency

We can show how latency decreases when we increase:

• number of LANs
• radius of LANs
• number of people
• how much people move/how much the stay in a LAN
• presence of Hub reachable from internet

Those graphs can have the mean latency in ticks on the y and the variable on the x. Maybe we can do some 3D graphs that shows how latency (y) decreases when we increase number of LANs (x) and increase number of people (z).

Redundancy of messages

Graph that shows the mean number of copies of messages (y) ...

• as time passes (time on x)
• after the simulation ends, with n people (x)
• after the simulation ends, with n LANs (x)

Number of hops between people

Graph that shows how the number of hops decreases as time passes. Probably this makes more sense on the transitive interests version.

[texdade]

I particularly like the idea of 3D graphs, I will have to recover some guides on matplotlib but it's definitely duable

[texdade]

The basic idea was to create multiple graphs out of the logs we have at our disposal:

• one 3D graph displaying the medium latency wrt the number of people and LANs on the simulation: this would be done by using an intermediate file like in the previous project, by using batch runs to gather data and appending to a csv file
• maybe it would be also interesting to display the number of follow/unfollow/block/unblock events created in time during a simulation, wrt to the number of people. This would be nice just because we'd have an additional kind of graph (and not only one kind like we did in the previous report)