This release fixes a bug that arose when using a default seed in mcmiser with numpy 1.9.0.
0.2
- MISER Monte Carlo, as proposed in Numerical Recipes. - All integration routines now respond to KeyboardInterrupt. - Added additional benchmarks for importance sampling.
0.1.5
generated, such that repeated calls to `mcquad` or `mcimport` that came within the same value of `int(time.time())` had the same seed.
Seeding is now left to the (more capable) hands of the RNG, which just gets passed `seed(None)` if the user hasn't specified a seed.