Skip to content

Hierarchical setups - allow RVs to be parameters? #58

@kevindavisross

Description

@kevindavisross

Example:

X = Uniform(0,1) Y = Binomial(10, X)
Every distribution will check to see whether each param is a number or an RV. If param is an RV, Y will be defined on the ProbabilitySpace whose .draw() method is:
def draw(): p = param.draw() return np.random.binomial(n=10, p=param)
Then modify the underlying ProbabilitySpace of X to also be on this probability space. (Question: does the above generate a probability space where the outcome is a pair (x, y)?)

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions