Package com.google.privacy.differentialprivacy

Interface Noise

  • All Known Implementing Classes:
    GaussianNoise, LaplaceNoise
    public interface Noise
    Interface for primitives that add noise to numerical data, for use in differential privacy operations.

    • Method Summary

      All Methods Static Methods Instance Methods Abstract Methods 
      Modifier and Type Method Description
      double addNoise​(double x, int l0Sensitivity, double lInfSensitivity, double epsilon, java.lang.Double delta)  
      long addNoise​(long x, int l0Sensitivity, long lInfSensitivity, double epsilon, java.lang.Double delta)  
      ConfidenceInterval computeConfidenceInterval​(double noisedX, int l0Sensitivity, double lInfSensitivity, double epsilon, java.lang.Double delta, double alpha)  
      ConfidenceInterval computeConfidenceInterval​(long noisedX, int l0Sensitivity, long lInfSensitivity, double epsilon, java.lang.Double delta, double alpha)  
      double computeQuantile​(double rank, double x, int l0Sensitivity, double lInfSensitivity, double epsilon, java.lang.Double delta)
      Calculates a value k s.t.
      static double getL1Sensitivity​(int l0Sensitivity, double lInfSensitivity)  
      static double getL2Sensitivity​(int l0Sensitivity, double lInfSensitivity)  
      com.google.differentialprivacy.SummaryOuterClass.MechanismType getMechanismType()  

    • Method Detail

      • addNoise

        double addNoise​(double x,
                int l0Sensitivity,
                double lInfSensitivity,
                double epsilon,
                @Nullable
                java.lang.Double delta)

      • addNoise

        long addNoise​(long x,
              int l0Sensitivity,
              long lInfSensitivity,
              double epsilon,
              @Nullable
              java.lang.Double delta)

      • computeConfidenceInterval

        ConfidenceInterval computeConfidenceInterval​(double noisedX,
                                             int l0Sensitivity,
                                             double lInfSensitivity,
                                             double epsilon,
                                             @Nullable
                                             java.lang.Double delta,
                                             double alpha)

      • computeConfidenceInterval

        ConfidenceInterval computeConfidenceInterval​(long noisedX,
                                             int l0Sensitivity,
                                             long lInfSensitivity,
                                             double epsilon,
                                             @Nullable
                                             java.lang.Double delta,
                                             double alpha)

      • getMechanismType

        com.google.differentialprivacy.SummaryOuterClass.MechanismType getMechanismType()

      • getL1Sensitivity

        static double getL1Sensitivity​(int l0Sensitivity,
                               double lInfSensitivity)

      • getL2Sensitivity

        static double getL2Sensitivity​(int l0Sensitivity,
                               double lInfSensitivity)

      • computeQuantile

        double computeQuantile​(double rank,
                       double x,
                       int l0Sensitivity,
                       double lInfSensitivity,
                       double epsilon,
                       @Nullable
                       java.lang.Double delta)
        Calculates a value k s.t. with probability rank the result of addNoise(double, int, double, double, java.lang.Double) with the given parameters will be less or equal to k.