原文：https://dzone.com/articles/shallow-retained-and-deep-size
If you ever worked with heap analysis, you might have encountered terms such as shallow, retained, and deep size of an object. It doesn’t really matter if we’re talking about analyzing a JVM heap dump with VisualVM or a JavaScript heap snapshot with Chrome DevTools – the terms are the same. The difference between shallow, retained, and deep size is rather straightforward and it can be easily depicted with the following images:
Shallow size – size of the object itself including only references to other objects, without objects which are referenced.
Retained size – size of the object including references to other objects and objects which are referenced, but are not shared with any other objects. In other words, the retained memory of an object is the amount of memory that would be freed if the object A was eligible to be collected by GC. Memory of the object E couldn’t be freed because it is also an instance variable of object B.
Deep size – size of the object including references to other objects and objects which are referenced, even if they are shared with other objects.