Hi
I'm trying to infer a weight matrix for a network. Every node's value in time t2 is dependent to it's value in time t1
If I replace SetTo with normal '=', and remove IfNot block, this code works. But I need the If block. Using SetTo, I get this error:
The left-hand side indices (outerRange) do not include the range 'index2', which appears on the right-hand side (perhaps implicitly by an open ForEach block).
It seems I can't use SetTo in a loop for a variable. Then how can I produce this weighted sum (Weighted sum of some values, in which weights are elements of a 2DArray).
I appreciate any help.
Thanks,
Zahra
* in weight matrix diagonal elements are weights Wii and are used for considering the effect of a node on itself at the next time step.
===============
VariableArray2D<double> w = Variable.Array<double>(outerRange, innerRange).Named("W");
/// here I initialize wij prior distributions.
var beta = Variable.Array<double>(outerRange).Named("beta");
beta[outerRange] = Variable.GaussianFromMeanAndVariance(1, 0.002).ForEach(outerRange);
using (ForEachBlock firstBlock = Variable.ForEach(outerRange))
{
Variable<double> selfEffect = -(w[outerRange,outerRange] * variableT1[outerRange]);
Variable<double> weightSum = Variable.New<double>();
using (ForEachBlock secondBlock = Variable.ForEach(innerRange))
{
var diagonalElement = Variable.Copy(secondBlock.Index == firstBlock.Index);
using (Variable.IfNot(diagonalElement))
{
Console.WriteLine("summing " + outerRange + " + " + innerRange);
weightSum.SetTo(weightSum
+ (w[outerRange, innerRange] * variableT1[innerRange]));
}
}
Variable<double> othersEffect = beta[outerRange] * Variable.Logistic(weightSum);
variableT2[outerRange] = selfEffect + othersEffect;
}
