Module icd 

Inverse CDF distribution builders.

Each dist_* helper builds a LutF64 containing the precomputed inverse CDF for a specific distribution. The DSL surface (#[polydat_node] form below) wraps each helper in a node that caches the LUT at construction via #[poly_const] and samples on every cycle.

This module also provides the UnitInterval and ClampF64 conversion nodes that are typically composed with LUT sampling in a DAG.

Supported Distributions

Continuous: Normal, Exponential, Uniform, Pareto, LogNormal, Weibull, Cauchy, Laplace, Beta, Gamma

Discrete: Zipf, Poisson, Binomial, Geometric

Structs

ClampF64

Clamp an f64 value to [min, max].

DistExponential

Sample from an exponential distribution Exp(rate).

DistNormal

Sample from a standard normal distribution N(mean, stddev).

DistPareto

Sample from a Pareto distribution Pareto(scale, shape).

DistUniform

Sample from a continuous uniform distribution U(min, max).

DistZipf

Sample from a Zipf distribution Zipf(n, exponent).

IcdExponential

Alias of dist_exponential.

IcdNormal

Alias of dist_normal. Preserved because workload examples and the host’s distribution binding both surface icd_normal as the public DSL name.

UnitInterval

Normalize a u64 to a uniform f64 in [0.0, 1.0).

Constants

DEFAULT_RESOLUTION

Default interpolation table resolution.

Functions

dist_beta_lut

Beta distribution: Beta(alpha, beta). Support: [0, 1].

dist_binomial_lut

Binomial distribution: Binomial(trials, p). Support: [0, trials].

dist_cauchy_lut

Cauchy distribution: Cauchy(location, scale). Support: (-∞, +∞).

dist_empirical_lut

Build a LUT from raw data points (continuous empirical distribution).

dist_empirical_weighted_lut

Build a LUT from weighted value-frequency pairs.

dist_exponential_lut

Exponential distribution: Exp(rate). Support: [0, +∞).

dist_gamma_lut

Gamma distribution: Gamma(shape, scale). Support: (0, +∞).

dist_geometric_lut

Geometric distribution: Geometric(p). Support: [1, +∞).

dist_laplace_lut

Laplace distribution: Laplace(location, scale). Support: (-∞, +∞).

dist_lognormal_lut

Log-normal distribution: LogN(mean, stddev). Support: (0, +∞).

dist_normal_lut

Normal distribution: N(mean, stddev).

dist_pareto_lut

Pareto distribution: Pareto(scale, shape). Support: [scale, +∞).

dist_poisson_lut

Poisson distribution: Poisson(lambda). Support: [0, +∞).

dist_uniform_lut

Uniform continuous distribution: U(min, max).

dist_weibull_lut

Weibull distribution: Weibull(shape, scale). Support: [0, +∞).

dist_zipf_lut

Zipf distribution: Zipf(n, exponent). Support: [1, n].