Class MeanSquaredError

  • All Implemented Interfaces:
    it.unibo.alchemist.boundary.Extractor

    
    public final class MeanSquaredError<T extends Object>
    extends AbstractDoubleExporter
                        

    Exports the Mean Squared Error for the concentration of some molecule, given another molecule that carries the correct result. The correct value is extracted from every node, then the provided {@link UnivariateStatistic} is applied to get a single, global correct value. Then, the actual value is extracted from every node, its value is compared (subtracted) to the computed correct value, it gets squared, and then logged.

    • Method Detail

      • extractData

         <T extends Any> Map<String, Double> extractData(Environment<T, ?> environment, Actionable<T> reaction, Time time, Long step)

        Extracts properties from an environment. The returned map must either:

        • contain a single element,

        • have the keys matching columnNames, or

        • be iterable in predictable order (namely, implement SortedMap or extend one of LinkedHashMap or ConcurrentLinkedHashMap).

        Parameters:
        environment -
        the {@link Environment}
        reaction -
        the last executed {@link Reaction}
        time -
        the current {@link Time}
        step -
        the simulation step
        Returns:

        the extracted properties with their names. The returned map must either: - contain a single element, - have the keys matching columnNames, or - be iterable in predictable order (namely, implement SortedMap or extend LinkedHashMap).