-
The test fails by selecting the wrong result ~3/100 times.
PRNG determinism in PECOptimizationFunction is done (in a hacky way) by replacing optuna's PRNG instance:
psyneulink/core/components/func…
-
```
The SMA PRNG is not working as expected.
Actually, the sma_random() function is never called due to bad address masking
in relevant functions.
I sent you a proposed fix by email, project owner.
…
-
```
The SMA PRNG is not working as expected.
Actually, the sma_random() function is never called due to bad address masking
in relevant functions.
I sent you a proposed fix by email, project owner.
…
-
We've already had a lot of discussion on this. Lets summarise algorithms here.
This is about small, fast PRNGs. Speed, size and performance in tests like PractRand and TestU01 is of interest here; …
-
```
The SMA PRNG is not working as expected.
Actually, the sma_random() function is never called due to bad address masking
in relevant functions.
I sent you a proposed fix by email, project owner.
…
-
Right now, when deciding the next ballot to sample, we use the PRNG used in the Haskell `random` library, seeded implicitly via use of the global RNG.
1. Decide if the PRNG in this library is accep…
-
See: https://github.com/MDAnalysis/mdanalysis/issues/1933
-
It would be cool to see more wipe method / prng added, because now we only have a few choice for prng and wiping method, for example you can try adding those wipe method:
Bruce Schneier: (7 passes)…
-
At the moment, there is not a good way to seed query language functions like `randomGaussian()` or `randomInt()` in a way that's thread-safe and efficient. This may take some real investigation.
-
- [ ] Remove num_bits from PRNG specification, lower level dialects closer to the implementation should support this
- [ ] Potentialy if random is a high level abstraction class, maybe we just need t…
asraa updated
1 month ago