Power Laws in Object Graphs

In here we present a collection of LOG-scale plots of the distributions of the number of references coming in and out of objects inside the program's memory. They tend to be lines, thus showing the presence of power law dependency in object graphs.

Incoming References LOG(# references) vs LOG(# objects)

Incoming References LOG(# references) vs LOG(# objects)

Outgoing References LOG(# references) vs LOG(# objects)

Outgoing References LOG(# references) vs LOG(# objects)

Incoming References LOG(# rank) vs LOG(# objects)

Incoming References LOG(# rank) vs LOG(# objects)

Outgoing References LOG(# rank) vs LOG(# objects)

Outgoing References LOG(# rank) vs LOG(# objects)

Incoming References LOG(# references) vs LOG(fraction of objects having more references)

Incoming References LOG(# references) vs LOG(fraction of objects having more references)

Static Incoming References LOG(# references) vs LOG(# objects)

Static Incoming References LOG(# references) vs LOG(# objects)

Incoming References vs Outgoing References

Incoming References vs Outgoing References together with the number of objects (white means many, black means few)

Incoming References vs Outgoing References (LOG-LOG Scale)

Incoming References vs Outgoing References (LOG-LOG Scale) together with the number of objects (white means many, black means few)