Opened 14 years ago
Closed 14 years ago
#348 closed enhancement (fixed)
Verify that psyco is being used to optimal advantage in ANUGA
| Reported by: | hudson | Owned by: | hudson |
|---|---|---|---|
| Priority: | normal | Milestone: | |
| Component: | Efficiency and optimisation | Version: | |
| Severity: | normal | Keywords: | |
| Cc: |
Description
It appears that psyco optimisation has been used selectively within ANUGA, and it would be worthwhile to check that it is being used in the right places, with the right settings.
Change History (1)
comment:1 Changed 14 years ago by hudson
- Resolution set to fixed
- Status changed from new to closed
Note: See
TracTickets for help on using
tickets.
I removed the selective psyco optimisations and just used the most aggressive form of compilation: psyco.full()
This compiles everything it can, and it resulted in ~20% speed increase on the benchmark and unit tests. There didn't seem to be a memory blowout.
Really, this should be done on the user side: we should leave it up to them if they want to use psyco or not. However, it's nice to have it running 20% faster out of the box. Does anyone know of any drawbacks of using aggressive psyco compilation?