Header menu logo bristlecone

Initialise Module

Functions and values

Function or value Description

initialise d rng

Full Usage: initialise d rng

Parameters:
Returns: TypedTensor<Vector, MeasureProduct<optim-space, MeasureOne>>

Generate a random point from bounds specified as a `Domain`. A value for each dimension is drawn from a univariate normal distribution, assuming that the bounds represent the 99th percentiles of the distribution.

d : Domain
rng : Random
Returns: TypedTensor<Vector, MeasureProduct<optim-space, MeasureOne>>

isInvalidTheta theta constraints

Full Usage: isInvalidTheta theta constraints

Parameters:
Returns: bool

Assesses if theta is valid based on the provided constraints.

theta : float<MeasureProduct<optim-space, MeasureOne>> seq
constraints : Constraint seq
Returns: bool

tryGenerateTheta f domain random n

Full Usage: tryGenerateTheta f domain random n

Parameters:
Returns: Result<Point, string>

Attempts to generate random theta based on starting bounds for n tries. If the theta does not meet given constraints, or `f` evaluates to NaN or an infinity then the algorithm tries again.

f : Objective
domain : Domain
random : Random
n : int
Returns: Result<Point, string>

Type something to start searching.