import distl
import numpy as np
g = distl.gaussian(10, 2)
distl obeys the seed set in np.random.seed
.
np.random.seed(1234)
g.sample()
8.255378595638142
np.random.seed(1234)
g.sample()
8.255378595638142
Alternatively, pass seed
to sample, plot, or plot_sample.
g.sample(seed=1234)
8.255378595638142
g.sample()
10.62204778579605
The random seed is respected for all distribution types, including Histogram.
h = g.to_histogram()
h.sample()
11.994789009229821
h.sample(seed=1234)
8.185549956549941
h.sample(seed=1234)
8.185549956549941
If you want a random seed to be used to multiple calls in the same execution, you can access a random array of integers via get_random_seed which can then be passed on to either np.random.seed or sample.
seed = distl.get_random_seed()
g.sample(seed=seed)
6.561067923451439
g.sample(seed=seed)
6.561067923451439
g.sample(seed=distl.get_random_seed())
11.27810575248316