locked
Engine Infer throwing error for Gaussian distribution RRS feed

  • Question

  • Here is my code:

      // Main Variables in the model
            public static VariableArray<double> workerAbility;
    
    
    
    
     public static void Infer(int[][] taskIndices, int[][] workerLabels)
            {
                int workerCount = workerLabels.Length;
                AttachData(taskIndices, workerLabels);
                Engine.NumberOfIterations = NumberOfIterations;
    
                workerAbility.AddAttribute(new Sequential());   // needed to get stable convergence
                taskDifficulty.AddAttribute(new Sequential());  // needed to get stable convergence
    
                // Get the compiled inference algorithm
                //var ca = Engine.GetCompiledInferenceAlgorithm(workerAbility, taskDifficulty, discrimination, trueLabel);
    
                workerAbilityPosterior = Engine.Infer<Gaussian[]>(workerAbility);//Error thrown here
                
            }

    While debugging, I found the error to be thrown here:

    Not quite sure what I might be doing wrong here? Where does that "value" in prob[value] come from? In the above code, Dimension=4, but "value" changes from 0 to 1 to 5. Not quite sure how this is happening and if what I am doing is actually the right way to infer Gaussian distribution ( I must be doing it wrong if there is error though haha)

    please advise


    • Edited by cindyak Tuesday, March 17, 2015 2:06 PM
    Tuesday, March 17, 2015 2:05 PM

All replies