penalized_bernoulli
Regularized Bernoulli marginal model.
PenalizedBernoulli(formula, *, penalties=(), device=None)
A Bernoulli marginal with coefficient penalties added to its training loss.
Parameters:
| Name | Type | Description | Default |
|---|---|---|---|
formula
|
dict of str to str, or str
|
Formula for the logit of the success probability. |
required |
penalties
|
sequence of CoefficientPenalty
|
See :mod: |
()
|
device
|
device or str
|
|
None
|
Examples:
>>> from scdesigner.distributions.penalties import CoefficientPenalty, QuadraticPenalty
>>> penalty = CoefficientPenalty("mean", ["x"], [1.0], QuadraticPenalty(lam=0.1))
>>> model = PenalizedBernoulli("~ x", penalties=[penalty])
Source code in scdesigner/src/scdesigner/distributions/penalized_bernoulli.py
9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 | |