Infer.NET

Variable Arrays in F#

Variable Arrays can be difficult to create and assign values to when dealing with large sets of complex data. This is especially true when 2-Dimensional and Jagged Arrays of data are involved. Therefore the module Variable provides various methods to make creation and assignment of such arrays both simpler and provide a more intuitive functional style of using such methods.

Creating Arrays using a generator function

The Variable.ArrayInit<'a> function provides an easy way of creating and assigning the values of a VariableArray given a generator function to provide the values for assignment. The type signature for this method is : Range -> (Range -> Variable<'a>) -> VariableArray<'a>. The following example shows how it can be used to create and initialise the mean and precision variableArrays used in the Mixture of Gaussians tutorial.

let means = Variable.ArrayInit  
              k (fun k -> Variable.VectorGaussianFromMeanAndPrecision(  
                            new Vector(2,0.0),  
                            PositiveDefiniteMatrix.IdentityScaledBy(2,0.01)))  
let precs = Variable.ArrayInit  
              k (fun k -> Variable.WishartFromShapeAndScale(  
                            100.0, PositiveDefiniteMatrix.IdentityScaledBy(2,0.01)))

Assigning arrays from a generator function

The Variable.AssignVariableArray function provides an easy way of assigning the values of a VariableArray given a generator function to provide the values for assignment. The type signature for this method is : VariableArray<'a> -> Range -> (Range -> Variable<'a>) -> VariableArray<'a>. The following example shows how it can be used to assign values to the variableArrays used in the Clinical trial tutorial using a generator function. Similarly, methods to assign values to 2-Dimensional and jagged VariableArrays are also provided in the form of Variable.AssignVariableArray2D with type signature : VariableArray2D<'a> -> Range -> (Range -> Variable<'a>) -> VariableArray2D<'a> and

let controlGroup = Variable.Observed<bool>([|false; false; true; false; false|])  
let treatedGroup = Variable.Observed<bool>([|true; false; true; true; true |])  
let i = controlGroup.Range  
let j = treatedGroup.Range  
// Prior on being an effective treatment  
let isEffective = Variable.Bernoulli(0.5).Named("isEffective");  
let probIfTreated = ref (Variable.New<float>())  
let probIfControl = ref (Variable.New<float>()) //rewrite If Block  
let f1 (vb1: Variable<bool>) =  
    probIfControl := Variable.Beta(1.0, 1.0).Named("probIfControl")  
    let controlGroup = Variable.AssignVariableArray controlGroup i (fun i ->Variable.Bernoulli(!probIfControl)) 
    let treatedGroup = Variable.AssignVariableArray treatedGroup j (fun j ->Variable.Bernoulli(!probIfTreated))  
    ()