- 28 Aug, 2020 4 commits
-
-
Noric Couderc authored
Uses a more efficient function to load CSV data.
-
Noric Couderc authored
Removed relying of map's mutability
-
Noric Couderc authored
The function uses the fact that rows in trace files are ordered to make more efficient loading of the traces.
-
Noric Couderc authored
uses flatmap instead of mapvalues and them toList
-
- 25 Aug, 2020 5 commits
-
-
Noric Couderc authored
Having a static DYNCLASSLOADER prevented some garbage collection, causing memory leaks.
-
Noric Couderc authored
Using the target_type of the metadata was wrong
-
Noric Couderc authored
Had a complicated way of getting the interface type of a collection class, now the InterfaceName enum type does the conversion. It is less general, but it's simpler.
-
Noric Couderc authored
Some benchmarks have a lot of duplicates, so we compress the data by aggregating them, the number of duplicates for each benchmark is then stored.
-
Noric Couderc authored
Some of the collections are not valid contestants for benchmarking because they don't have the same semantics.
-
- 21 Aug, 2020 8 commits
-
-
Noric Couderc authored
-
Noric Couderc authored
Error messages were displayed when running the test
-
Noric Couderc authored
-
Noric Couderc authored
-
Noric Couderc authored
Thought this might help with performance (not sure it does)
-
Noric Couderc authored
When two traces are from the same allocation site, with the same target type and the exact same list of methods, we only keep one.
-
Noric Couderc authored
We used groupby and then processed each group, so we can clear as we go.
-
Noric Couderc authored
The benchmarks didn't use the numberSamples and numberWarmups given to the constructor
-
- 20 Aug, 2020 4 commits
-
-
Noric Couderc authored
Relied on streams more instead of lists, trying to get faster running of generated benchmarks Removed double-generation of benchmarks
-
Noric Couderc authored
There is probably a better way, since some options are then invalid
-
Noric Couderc authored
The benchmark part was too unstable...
-
Noric Couderc authored
Each traces is loaded, grouped, and benchmarks are created to try each collection for each benchmarks, then suggestions are created
-
- 19 Aug, 2020 14 commits
-
-
Noric Couderc authored
Originally, when loading a trace, only one benchmark was created but there's now a function to also generate benchmarks so that we can run them and make comparisons
-
Noric Couderc authored
Changed the expected values to fit the new format for benchmark identifiers
-
Noric Couderc authored
List.of() doesn't exist for java 8
-
Noric Couderc authored
-
Noric Couderc authored
-
Noric Couderc authored
Theres a test for running a benchmark based on a trace and one from trace benchmarks loaded from a CSV data
-
Noric Couderc authored
The user is able to change the number of iterations for measurement and warmup in SyntheticBenchmarkRunner
-
Noric Couderc authored
-
Noric Couderc authored
A trace with a different object ID and location should have a different ID.
-
Noric Couderc authored
-
Noric Couderc authored
-
Noric Couderc authored
-
Noric Couderc authored
We check that the ids are indeed different if we generate two benchmarks correctly
-
Noric Couderc authored
-
- 18 Aug, 2020 5 commits
-
-
Noric Couderc authored
Removed double iteration over all benchmarks to build the list of benchmark indexes. Also fixed a broken switch statement.
-
Noric Couderc authored
Was necessary to enable serialization
-
Noric Couderc authored
The test writes benchmarks to memory and then reads them
-
Noric Couderc authored
Means that deserializing from memory is also possible
-
Noric Couderc authored
-