Package com.google.privacy.differentialprivacy

Class BoundedQuantiles.Params.Builder

java.lang.Object

com.google.privacy.differentialprivacy.BoundedQuantiles.Params.Builder

Enclosing class:

BoundedQuantiles.Params

public abstract static class BoundedQuantiles.Params.Builder
extends java.lang.Object

Constructor Summary

Constructors

Constructor	Description
Builder()

Method Summary

Modifier and Type	Method	Description
abstract BoundedQuantiles.Params.Builder	branchingFactor(int value)	The number of children for every non-leaf node of the quantile tree used to store the entries.
BoundedQuantiles	build()
abstract BoundedQuantiles.Params.Builder	delta(java.lang.Double value)	Delta DP parameter.
abstract BoundedQuantiles.Params.Builder	epsilon(double value)	Epsilon DP parameter.
abstract BoundedQuantiles.Params.Builder	lower(double value)	Lower bound for the entries added to the distribution.
abstract BoundedQuantiles.Params.Builder	maxContributionsPerPartition(int value)	Max contributions per partition from a single privacy unit (e.g., an individual).
abstract BoundedQuantiles.Params.Builder	maxPartitionsContributed(int value)	Maximum number of partitions that a single privacy unit (e.g., an individual) is allowed to contribute to.
abstract BoundedQuantiles.Params.Builder	noise(Noise value)	Noise that will be used to make the quantiles differentially private.
abstract BoundedQuantiles.Params.Builder	treeHeight(int value)	The height of the quantile tree used to store the entries.
abstract BoundedQuantiles.Params.Builder	upper(double value)	Upper bound for the entries added to the distribution.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

Builder

public Builder()

Method Detail

epsilon

public abstract BoundedQuantiles.Params.Builder epsilon(double value)

Epsilon DP parameter.

delta

public abstract BoundedQuantiles.Params.Builder delta(@Nullable
                                                      java.lang.Double value)

Delta DP parameter.

Note that Laplace noise does not use delta. Hence, delta should not be set when Laplace noise is used.

maxPartitionsContributed

public abstract BoundedQuantiles.Params.Builder maxPartitionsContributed(int value)

Maximum number of partitions that a single privacy unit (e.g., an individual) is allowed to contribute to.

maxContributionsPerPartition

public abstract BoundedQuantiles.Params.Builder maxContributionsPerPartition(int value)

Max contributions per partition from a single privacy unit (e.g., an individual).

noise

public abstract BoundedQuantiles.Params.Builder noise(Noise value)

Noise that will be used to make the quantiles differentially private.

lower

public abstract BoundedQuantiles.Params.Builder lower(double value)

Lower bound for the entries added to the distribution. Any entires smaller than this value will be set to this value.

upper

public abstract BoundedQuantiles.Params.Builder upper(double value)

Upper bound for the entries added to the distribution. Any entires greater than this value will be set to this value.

treeHeight

public abstract BoundedQuantiles.Params.Builder treeHeight(int value)

The height of the quantile tree used to store the entries. Should be at least 1. This parameter is not public, it should be used only by other aggregation functions inside the library.

branchingFactor

public abstract BoundedQuantiles.Params.Builder branchingFactor(int value)

The number of children for every non-leaf node of the quantile tree used to store the entries. Should be at least 2. This parameters is not public, it should be used only by other aggregation functions inside the library.

build

public BoundedQuantiles build()