Fan-in is a measure of the number of functions or methods that call some other function or method (say X). Fan-out is the number of functions that are called by function X. It is used as a call graph based metrics. My question is how to measure fan-in/out from a call graph and where we can use this metrics?
A sample call graph is given as follows.
I haven't work in this area for a long while, but can at least get the discussion going. The measurement of the fan out/in of a given node is simply the count of "arrows" in and out. For example "process" in your example chart has a fan out of three. The summary measures for the entire chart are based on all the nodes fan in/out and would include mean, min, max, and median. The fan out is a reasonable start at or component of a procedure complexity measure. The fan in can be useful as a screening measure of the criticality of the procedure and the importance of its execution speed/efficiency.
You should have a look at centrality metrics. Specifically indegree and outdegree centrality
http://en.wikipedia.org/wiki/Centrality#Degree_centrality
Thanks Maya Daneva for your reply.
I am trying to calculate fan in-out to measure dependency between functions. I am wondering if it is possible to evaluate my testing technique with this.
- Badges
- Science topic