- (TZ) Version bumped to 1.1.0
- (TZ) tinspin-indexes 2.0.0
- (TZ) Changed data type of indexes from Singleton Object (or sometimes double[]) to Integer ID.
- (TZ) Added support for multimap tests
- (TZ) Updated README with proper description of output format
- (TZ) Modularization, new base module tinspin-common
- (TZ) Added new Example
- (TZ) Added (very basic!) import for HDF5 files.
- (TZ) Updated Window Query generator again
- (TZ) Re-added repeated testing of update()
- (TZ) Use result count only from first run, for comparability
- (TZ) Updated timings to use nanoSecs internally
- (TZ) CHANGE: WQ timings now return time/query instead of time/result
- (TZ) New logging output with ops/sec instead of time/op
- (TZ) Operations without side effects (window query, point query, kNN search) are repeatedly executed until a minimum time has passed (in order to avoid problems with warm-up)
- (TZ) Fixed window query generator for dim>40
- (TZ) Updated reference to tinspin-indexes to 1.5.1
- (TZ) Updated reference to tinspin-indexes to 1.4.0
- (TZ) Added Wrapper for TinSpin-Indexes KD-Tree
- (TZ) Added CLUSTER Gauss generators, cleaned up update distance, general cleanup
- (TZ) Refactored random number generation to be reinitialized for each task.