pymc.logcdf#

pymc.logcdf(rv, value, warn_rvs=None, **kwargs)[source]#

Create a graph for the log-CDF of a random variable.

Parameters:

rvTensorVariable
valuetensor_like: Should be the same type (shape and dtype) as the rv.
warn_rvsbool, default True: Warn if RVs were found in the logcdf graph. This can happen when a variable has other random variables as inputs. In that case, those random variables should be replaced by their respective values.

Returns:

logpTensorVariable

Raises:

RuntimeError: If the logcdf cannot be derived.

Examples

Create a compiled function that evaluates the logcdf of a variable

import pymc as pm
import pytensor.tensor as pt

mu = pt.scalar("mu")
rv = pm.Normal.dist(mu, 1.0)

value = pt.scalar("value")
rv_logcdf = pm.logcdf(rv, value)

# Use .eval() for debugging
print(rv_logcdf.eval({value: 0.9, mu: 0.0}))  # -0.2034146

# Compile a function for repeated evaluations
rv_logcdf_fn = pm.compile_pymc([value, mu], rv_logcdf)
print(rv_logcdf_fn(value=0.9, mu=0.0))  # -0.2034146

Derive the graph for a transformation of a RandomVariable

import pymc as pm
import pytensor.tensor as pt

mu = pt.scalar("mu")
rv = pm.Normal.dist(mu, 1.0)
exp_rv = pt.exp(rv)

value = pt.scalar("value")
exp_rv_logcdf = pm.logcdf(exp_rv, value)

# Use .eval() for debugging
print(exp_rv_logcdf.eval({value: 0.9, mu: 0.0}))  # -0.78078813

# Compile a function for repeated evaluations
exp_rv_logcdf_fn = pm.compile_pymc([value, mu], exp_rv_logcdf)
print(exp_rv_logcdf_fn(value=0.9, mu=0.0))  # -0.78078813

Define a CustomDist logcdf

import pymc as pm
import pytensor.tensor as pt

def normal_logcdf(value, mu, sigma):
    return pm.logp(pm.Normal.dist(mu, sigma), value)

with pm.Model() as model:
    mu = pm.Normal("mu")
    sigma = pm.HalfNormal("sigma")
    pm.CustomDist("x", mu, sigma, logcdf=normal_logcdf)