Module opendp::measurements::alp

Constants

• ALPHA_DEFAULT 🔒
• SIZE_FACTOR_DEFAULT 🔒

Functions

• are_parameters_invalid 🔒
  Reject any choice of scale or alpha such that scale / alpha < 2^-52
• compute_estimate 🔒
  Estimates the value of an entry based on its noisy bit representation.
• compute_prob 🔒
• compute_projection 🔒
  Computes the DP projection.
• estimate_unary 🔒
  Estimate the value of an entry based on its noisy bit representation.
• exponent_next_power_of_two 🔒
  Computes ceil(log_2(x))
• hash 🔒
  A hash function with type: [2^64] -> [2^l]
• make_alp_queryable
  Measurement to release a queryable containing a DP projection of bounded sparse data.
• make_alp_state
  Measurement to compute a DP projection of bounded sparse data.
• make_alp_state_with_hashers
  Measurement to compute a DP projection of bounded sparse data.
• post_alp_state_to_queryable
  Make a postprocessor that wraps the AlpState in a Queryable object.
• pre_hash 🔒
  Hash a key of type K into a u64
• sample_hash_function 🔒
  Samples a random hash function with type: K -> [2^l]
• scale_and_round 🔒

Type Definitions

• BitVector 🔒
• HashFunction 🔒
• SparseDomain 🔒
  Implementation of a mechanism for representing sparse integer queries e.g. a sparse histogram. The mechanism was introduced in the paper: