Package index • R Documentation

Accuracy

The accuracy module provides functions for converting between accuracy and scale parameters.

accuracy_to_discrete_gaussian_scale(): Convert a desired accuracy (tolerance) into a discrete gaussian noise scale at a statistical significance level alpha.
accuracy_to_discrete_laplacian_scale(): Convert a desired accuracy (tolerance) into a discrete Laplacian noise scale at a statistical significance level alpha.
accuracy_to_gaussian_scale(): Convert a desired accuracy (tolerance) into a gaussian noise scale at a statistical significance level alpha.
accuracy_to_laplacian_scale(): Convert a desired accuracy (tolerance) into a Laplacian noise scale at a statistical significance level alpha.
discrete_gaussian_scale_to_accuracy(): Convert a discrete gaussian scale into an accuracy estimate (tolerance) at a statistical significance level alpha.
discrete_laplacian_scale_to_accuracy(): Convert a discrete Laplacian scale into an accuracy estimate (tolerance) at a statistical significance level alpha.
gaussian_scale_to_accuracy(): Convert a gaussian scale into an accuracy estimate (tolerance) at a statistical significance level alpha.
laplacian_scale_to_accuracy(): Convert a Laplacian scale into an accuracy estimate (tolerance) at a statistical significance level alpha.

Combinators

The combinators module provides functions for combining transformations and measurements.

make_adaptive_composition(): adaptive composition constructor
make_approximate(): approximate constructor
make_basic_composition(): basic composition constructor
make_chain_mt(): chain mt constructor
make_chain_pm(): chain pm constructor
make_chain_tt(): chain tt constructor
make_composition(): composition constructor
make_fix_delta(): fix delta constructor
make_fixed_approxDP_to_approxDP(): fixed approxDP to approxDP constructor
make_fully_adaptive_composition(): fully adaptive composition constructor
make_population_amplification(): population amplification constructor
make_privacy_filter(): privacy filter constructor
make_pureDP_to_zCDP(): pureDP to zCDP constructor
make_select_private_candidate(): select private candidate constructor
make_sequential_composition(): sequential composition constructor
make_zCDP_to_approxDP(): zCDP to approxDP constructor
then_adaptive_composition(): partial adaptive composition constructor
then_approximate(): partial approximate constructor
then_basic_composition(): partial basic composition constructor
then_chain_mt(): partial chain mt constructor
then_chain_pm(): partial chain pm constructor
then_chain_tt(): partial chain tt constructor
then_composition(): partial composition constructor
then_fix_delta(): partial fix delta constructor
then_fixed_approxDP_to_approxDP(): partial fixed approxDP to approxDP constructor
then_fully_adaptive_composition(): partial fully adaptive composition constructor
then_population_amplification(): partial population amplification constructor
then_privacy_filter(): partial privacy filter constructor
then_pureDP_to_zCDP(): partial pureDP to zCDP constructor
then_select_private_candidate(): partial select private candidate constructor
then_sequential_composition(): partial sequential composition constructor
then_zCDP_to_approxDP(): partial zCDP to approxDP constructor

Core

The core module provides functions for accessing the fields of transformations and measurements.

function_eval(): Eval the function with arg.
measurement_check(): Check the privacy relation of the measurement at the given d_in, d_out
measurement_function(): Get the function from a measurement.
measurement_input_carrier_type(): Get the input (carrier) data type of this.
measurement_input_distance_type(): Get the input distance type of measurement.
measurement_input_domain(): Get the input domain from a measurement.
measurement_input_metric(): Get the input domain from a measurement.
measurement_invoke(): Invoke the measurement with arg. Returns a differentially private release.
measurement_map(): Use the measurement to map a given d_in to d_out.
measurement_output_distance_type(): Get the output distance type of measurement.
measurement_output_measure(): Get the output domain from a measurement.
new_function(): Construct a Function from a user-defined callback. Can be used to build a postprocessor.
new_queryable(): Construct a queryable from a user-defined transition function.
odometer_input_carrier_type(): Get the input (carrier) data type of this.
odometer_input_domain(): Get the input domain from a odometer.
odometer_input_metric(): Get the input domain from a odometer.
odometer_invoke(): Invoke the odometer with arg. Returns a differentially private release.
odometer_output_measure(): Get the output domain from a odometer.
odometer_queryable_invoke(): Eval the odometer queryable with an invoke query.
odometer_queryable_invoke_type(): Get the invoke query type of an odometer queryable.
odometer_queryable_privacy_loss(): Retrieve the privacy loss of an odometer queryable.
odometer_queryable_privacy_loss_type(): Get the map query type of an odometer queryable.
queryable_eval(): Eval the queryable with query. Returns a differentially private release.
queryable_query_type(): Get the query type of queryable.
transformation_check(): Check the privacy relation of the measurement at the given d_in, d_out
transformation_function(): Get the function from a transformation.
transformation_input_carrier_type(): Get the input (carrier) data type of this.
transformation_input_distance_type(): Get the input distance type of transformation.
transformation_input_domain(): Get the input domain from a transformation.
transformation_input_metric(): Get the input domain from a transformation.
transformation_invoke(): Invoke the transformation with arg. Returns a differentially private release.
transformation_map(): Use the transformation to map a given d_in to d_out.
transformation_output_distance_type(): Get the output distance type of transformation.
transformation_output_domain(): Get the output domain from a transformation.
transformation_output_metric(): Get the output domain from a transformation.

Domains

The domains module provides functions for creating and using domains.

atom_domain(): Construct an instance of AtomDomain.
bitvector_domain(): Construct an instance of BitVectorDomain.
domain_carrier_type(): Get the carrier type of a domain.
domain_debug(): Debug a domain.
domain_type(): Get the type of a domain.
map_domain(): Construct an instance of MapDomain.
option_domain(): Construct an instance of OptionDomain.
user_domain(): Construct a new UserDomain. Any two instances of an UserDomain are equal if their string descriptors are equal. Contains a function used to check if any value is a member of the domain.
vector_domain(): Construct an instance of VectorDomain.

Measurements

The measurements module provides functions that apply calibrated noise to data to ensure differential privacy.

make_alp_queryable(): alp queryable constructor
make_canonical_noise(): canonical noise constructor
make_gaussian(): gaussian constructor
make_gaussian_threshold(): gaussian threshold constructor
make_geometric(): geometric constructor
make_laplace(): laplace constructor
make_laplace_threshold(): laplace threshold constructor
make_noise(): noise constructor
make_noise_threshold(): noise threshold constructor
make_noisy_max(): noisy max constructor
make_noisy_top_k(): noisy top k constructor
make_private_quantile(): private quantile constructor
make_randomized_response(): randomized response constructor
make_randomized_response_bitvec(): randomized response bitvec constructor
make_randomized_response_bool(): randomized response bool constructor
make_report_noisy_max_gumbel(): report noisy max gumbel constructor
make_user_measurement(): user measurement constructor
then_alp_queryable(): partial alp queryable constructor
then_canonical_noise(): partial canonical noise constructor
then_gaussian(): partial gaussian constructor
then_gaussian_threshold(): partial gaussian threshold constructor
then_geometric(): partial geometric constructor
then_laplace(): partial laplace constructor
then_laplace_threshold(): partial laplace threshold constructor
then_noise(): partial noise constructor
then_noise_threshold(): partial noise threshold constructor
then_noisy_max(): partial noisy max constructor
then_noisy_top_k(): partial noisy top k constructor
then_private_quantile(): partial private quantile constructor
then_randomized_response(): partial randomized response constructor
then_randomized_response_bitvec(): partial randomized response bitvec constructor
then_randomized_response_bool(): partial randomized response bool constructor
then_report_noisy_max_gumbel(): partial report noisy max gumbel constructor
then_user_measurement(): partial user measurement constructor

Measures

The measures modules provides functions that measure the distance between probability distributions.

approximate(): Privacy measure used to define \(\delta\)-approximate PM-differential privacy.
fixed_smoothed_max_divergence(): Privacy measure used to define \((\epsilon, \delta)\)-approximate differential privacy.
max_divergence(): Privacy measure used to define \(\epsilon\)-pure differential privacy.
measure_debug(): Debug a measure.
measure_distance_type(): Get the distance type of a measure.
measure_type(): Get the type of a measure.
new_privacy_profile(): Construct a PrivacyProfile from a user-defined callback.
renyi_divergence(): Privacy measure used to define \(\epsilon(\alpha)\)-Rényi differential privacy.
smoothed_max_divergence(): Privacy measure used to define \(\epsilon(\delta)\)-approximate differential privacy.
user_divergence(): Privacy measure with meaning defined by an OpenDP Library user (you).
zero_concentrated_divergence(): Privacy measure used to define \(\rho\)-zero concentrated differential privacy.

Metrics

The metrics module provides functions that measure the distance between two elements of a domain.

absolute_distance(): Construct an instance of the AbsoluteDistance metric.
change_one_distance(): Construct an instance of the ChangeOneDistance metric.
discrete_distance(): Construct an instance of the DiscreteDistance metric.
hamming_distance(): Construct an instance of the HammingDistance metric.
insert_delete_distance(): Construct an instance of the InsertDeleteDistance metric.
l01inf_distance(): Construct an instance of the L01InfDistance metric.
l02inf_distance(): Construct an instance of the L02InfDistance metric.
l1_distance(): Construct an instance of the L1Distance metric.
l2_distance(): Construct an instance of the L2Distance metric.
linf_distance(): Construct an instance of the LInfDistance metric.
metric_debug(): Debug a metric.
metric_distance_type(): Get the distance type of a metric.
metric_type(): Get the type of a metric.
symmetric_distance(): Construct an instance of the SymmetricDistance metric.
user_distance(): Construct a new UserDistance. Any two instances of an UserDistance are equal if their string descriptors are equal.

Mod

The mod module provides the classes which implement the OpenDP Programming Framework, as well as utilities for enabling features and finding parameter values.

binary_search(): Find the closest passing value to the decision boundary of predicate
binary_search_chain(): Find the highest-utility (d_in, d_out)-close Transformation or Measurement.
binary_search_param(): Solve for the ideal constructor argument to make_chain
disable_features(): Disable features in the opendp package.
enable_features(): Enable features for the opendp package.
hashitems(): extract heterogeneously typed keys and values from a hashtab
new_domain(): new domain
new_function_internal(): new function
new_hashtab(): create an instance of a hashtab from keys and values
new_measure(): new measure
new_measurement(): new measurement
new_metric(): new metric
new_odometer(): new odometer
new_odometer_queryable_internal(): new odometer queryable
new_privacy_profile_internal(): new privacy profile
new_queryable_internal(): new queryable
new_transformation(): new transformation
to_str(<default>): Convert a format-able value to a string representation
to_str(<hashtab>): Convert hashtab to a string representation

Transformations

The transformations module provides functions that deterministically transform datasets.

choose_branching_factor(): Returns an approximation to the ideal branching_factor for a dataset of a given size, that minimizes error in cdf and quantile estimates based on b-ary trees.
make_b_ary_tree(): b ary tree constructor
make_bounded_float_checked_sum(): bounded float checked sum constructor
make_bounded_float_ordered_sum(): bounded float ordered sum constructor
make_bounded_int_monotonic_sum(): bounded int monotonic sum constructor
make_bounded_int_ordered_sum(): bounded int ordered sum constructor
make_bounded_int_split_sum(): bounded int split sum constructor
make_cast(): cast constructor
make_cast_default(): cast default constructor
make_cast_inherent(): cast inherent constructor
make_cdf(): cdf constructor
make_clamp(): clamp constructor
make_consistent_b_ary_tree(): consistent b ary tree constructor
make_count(): count constructor
make_count_by(): count by constructor
make_count_by_categories(): count by categories constructor
make_count_distinct(): count distinct constructor
make_create_dataframe(): create dataframe constructor
make_df_cast_default(): df cast default constructor
make_df_is_equal(): df is equal constructor
make_drop_null(): drop null constructor
make_find(): find constructor
make_find_bin(): find bin constructor
make_identity(): identity constructor
make_impute_constant(): impute constant constructor
make_impute_uniform_float(): impute uniform float constructor
make_index(): index constructor
make_is_equal(): is equal constructor
make_is_null(): is null constructor
make_lipschitz_float_mul(): lipschitz float mul constructor
make_mean(): mean constructor
make_metric_bounded(): metric bounded constructor
make_metric_unbounded(): metric unbounded constructor
make_ordered_random(): ordered random constructor
make_quantile_score_candidates(): quantile score candidates constructor
make_quantiles_from_counts(): quantiles from counts constructor
make_resize(): resize constructor
make_select_column(): select column constructor
make_sized_bounded_float_checked_sum(): sized bounded float checked sum constructor
make_sized_bounded_float_ordered_sum(): sized bounded float ordered sum constructor
make_sized_bounded_int_checked_sum(): sized bounded int checked sum constructor
make_sized_bounded_int_monotonic_sum(): sized bounded int monotonic sum constructor
make_sized_bounded_int_ordered_sum(): sized bounded int ordered sum constructor
make_sized_bounded_int_split_sum(): sized bounded int split sum constructor
make_split_dataframe(): split dataframe constructor
make_split_lines(): split lines constructor
make_split_records(): split records constructor
make_subset_by(): subset by constructor
make_sum(): sum constructor
make_sum_of_squared_deviations(): sum of squared deviations constructor
make_unordered(): unordered constructor
make_user_transformation(): user transformation constructor
make_variance(): variance constructor
then_b_ary_tree(): partial b ary tree constructor
then_bounded_float_checked_sum(): partial bounded float checked sum constructor
then_bounded_float_ordered_sum(): partial bounded float ordered sum constructor
then_bounded_int_monotonic_sum(): partial bounded int monotonic sum constructor
then_bounded_int_ordered_sum(): partial bounded int ordered sum constructor
then_bounded_int_split_sum(): partial bounded int split sum constructor
then_cast(): partial cast constructor
then_cast_default(): partial cast default constructor
then_cast_inherent(): partial cast inherent constructor
then_cdf(): partial cdf constructor
then_clamp(): partial clamp constructor
then_consistent_b_ary_tree(): partial consistent b ary tree constructor
then_count(): partial count constructor
then_count_by(): partial count by constructor
then_count_by_categories(): partial count by categories constructor
then_count_distinct(): partial count distinct constructor
then_create_dataframe(): partial create dataframe constructor
then_df_cast_default(): partial df cast default constructor
then_df_is_equal(): partial df is equal constructor
then_drop_null(): partial drop null constructor
then_find(): partial find constructor
then_find_bin(): partial find bin constructor
then_identity(): partial identity constructor
then_impute_constant(): partial impute constant constructor
then_impute_uniform_float(): partial impute uniform float constructor
then_index(): partial index constructor
then_is_equal(): partial is equal constructor
then_is_null(): partial is null constructor
then_lipschitz_float_mul(): partial lipschitz float mul constructor
then_mean(): partial mean constructor
then_metric_bounded(): partial metric bounded constructor
then_metric_unbounded(): partial metric unbounded constructor
then_ordered_random(): partial ordered random constructor
then_quantile_score_candidates(): partial quantile score candidates constructor
then_quantiles_from_counts(): partial quantiles from counts constructor
then_resize(): partial resize constructor
then_select_column(): partial select column constructor
then_sized_bounded_float_checked_sum(): partial sized bounded float checked sum constructor
then_sized_bounded_float_ordered_sum(): partial sized bounded float ordered sum constructor
then_sized_bounded_int_checked_sum(): partial sized bounded int checked sum constructor
then_sized_bounded_int_monotonic_sum(): partial sized bounded int monotonic sum constructor
then_sized_bounded_int_ordered_sum(): partial sized bounded int ordered sum constructor
then_sized_bounded_int_split_sum(): partial sized bounded int split sum constructor
then_split_dataframe(): partial split dataframe constructor
then_split_lines(): partial split lines constructor
then_split_records(): partial split records constructor
then_subset_by(): partial subset by constructor
then_sum(): partial sum constructor
then_sum_of_squared_deviations(): partial sum of squared deviations constructor
then_unordered(): partial unordered constructor
then_user_transformation(): partial user transformation constructor
then_variance(): partial variance constructor

Typing

The typing module provides utilities that bridge between R and Rust types. OpenDP relies on precise descriptions of data types to make its security guarantees: These are more natural in Rust with its fine-grained type system, but they may feel out of place in R. These utilities try to fill that gap.

BitVector: type signature for a BitVector
ExtrinsicObject: type signature for an arbitrary R object preserved across FFI
String: type signature for a string
bool: type signature for a boolean
f32: type signature for a 32-bit floating point number
f64: type signature for a 64-bit floating point number
i128: type signature for a 128-bit signed integer
i16: type signature for a 16-bit signed integer
i32: type signature for a 32-bit signed integer
i64: type signature for a 64-bit signed integer
i8: type signature for an 8-bit signed integer
parse_or_infer(): Parse a runtime type or infer it from an example
rt_infer(): Infer a runtime type from a public example
rt_parse(): Parse a runtime type descriptor into a runtime_type object
u128: type signature for a 128-bit unsigned integer
u16: type signature for a 16-bit unsigned integer
u32: type signature for a 32-bit unsigned integer
u64: type signature for a 64-bit unsigned integer
u8: type signature for an 8-bit unsigned integer
usize: type signature for a pointer-sized unsigned integer

Other

This should be empty if correctly configured. Please file an issue if any functions are listed here.

then_postprocess(): Compose a measurement with a postprocessing function.
to_then(): Convert a make_ constructor into a then_ constructor.