0.8 Update

This changes does not change functionality and will not break any implementation. All changes are internal only based on recieved feedback.

Optimisation

• Removed unnecessary updateIndex() call with an object is added, when we can easily determine the new entries' index value

Improved Benchmarking

• Remove loops and just use nano seconds
• Fix JVM memory allocation
• Setup & Teardown collections during each benchmark iteration
• Removed pointless map.addAll bench
• Removed unfair consume operation from mapAddSingle

0.7 Update

Implemented clear() method

This change only adds functionality and will not break any previous implementations.

0.6 Update

This changes does not change functionality and will not break any
implementation. All changes are internal only based on recieved feedback.

Changes to Collection Class:

1. Updated contains to just use collection.stream.anyMatch
2. Added parameter guards to protect against invalid elements
3. get() will throw an illegal state exception if the collection is
   empty
4. get() asserts object is not null (Which if the collection contains
   elements, it never should be).
5. JavaDoc has been updated to document these changes

Changes to Tests:

1. Added new tests to assert all errors are thrown correctly
2. Added new test to assert get() never returns null in extreme edge
   cases
3. Increased repeated test amounts
4. Updated int comparison to use assertEquals not assertTrue x == y

Enjoy!
ProbabilityCollection.zip

Initial Release

Initial Release

Even though the data-structure has been thoroughly tested, it has not been used in a real production environment. I plan to increment the version slowly if bugs are encountered until finally we can reach a 1.0 release. If no bugs are uncovered, then over time the version will be bumped to 1.0 anyway.

Please find the v0.5 java file attached and feel free to post issues / questions,
Lewys Davies

ProbabilityCollection.zip