nessai.evidence

Functions related to computing the evidence.

Functions

logsubexp(x, y)	Helper function to compute the exponential
log_integrate_log_trap(log_func, log_support)	Trapezoidal integration of given log(func). Returns log of the integral.
log_evidence_from_ins_samples(→ float)	Compute the log-evidence for a given set of samples from the importance

Module Contents

nessai.evidence.logsubexp(x, y)

Helper function to compute the exponential of a difference between two numbers

Computes: x + np.log1p(-np.exp(y-x))

Parameters:

x, yfloat or array_like

Inputs

nessai.evidence.log_integrate_log_trap(log_func, log_support)

Trapezoidal integration of given log(func). Returns log of the integral.

Parameters:

log_funcarray_like

Log values of the function to integrate over.

log_supportarray_like

Log prior-volumes for each value.

Returns:

float

Log of the result of the integral.

nessai.evidence.log_evidence_from_ins_samples(samples: numpy.ndarray) → float

Compute the log-evidence for a given set of samples from the importance nested sampler.

Parameters:

samplesnumpy.ndarray

Array of samples from the importance nested samples. Must have the logW field.

Returns:

float

The log-evidence