scvi.distributions.NegativeBinomialMixture

class scvi.distributions.NegativeBinomialMixture(mu1, mu2, theta1, mixture_logits, theta2=None, validate_args=False)

Bases: Distribution

Negative binomial mixture distribution.

See NegativeBinomial for further description of parameters.

Parameters:

• mu1 (Tensor) – Mean of the component 1 distribution.

• mu2 (Tensor) – Mean of the component 2 distribution.

• theta1 (Tensor) – Inverse dispersion for component 1.

• mixture_logits (Tensor) – Logits scale probability of belonging to component 1.

• theta2 (Tensor (default: None)) – Inverse dispersion for component 1. If None, assumed to be equal to theta1.

• validate_args (bool (default: False)) – Raise ValueError if arguments do not match constraints

Attributes table

arg_constraints
mean	Returns the mean of the distribution.
mixture_probs
support

Methods table

get_normalized(key)
log_prob(value)	Log probability.
sample([sample_shape])	Sample from the distribution.

Attributes

NegativeBinomialMixture.arg_constraints = {'mixture_logits': Real(), 'mixture_probs': HalfOpenInterval(lower_bound=0.0, upper_bound=1.0), 'mu1': GreaterThanEq(lower_bound=0), 'mu2': GreaterThanEq(lower_bound=0), 'theta1': GreaterThanEq(lower_bound=0)}

NegativeBinomialMixture.mean

NegativeBinomialMixture.mixture_probs

NegativeBinomialMixture.support = IntegerGreaterThan(lower_bound=0)

Methods

NegativeBinomialMixture.get_normalized(key)

Return type:

Tensor

NegativeBinomialMixture.log_prob(value)

Log probability.

Return type:

Tensor

NegativeBinomialMixture.sample(sample_shape=None)

Sample from the distribution.

Return type:

Tensor