Enum Distribution

pub enum Distribution {
    Uniform(UniformDistribution),
    Exponential(ExponentialDistribution),
    Gaussian(GaussianDistribution),
    Bernoulli(BernoulliDistribution),
    Generic(GenericDistribution),
}

This object defines probabilistic distributions that encode uncertain information about a single, scalar value. Currently, we support five core statistical distributions. New variants will be added over time.

This object is the lowest-level object in the statistics hierarchy, and it is the main unit of calculus when evaluating expressions in a statistical context. Notions like column and table statistics are built on top of this object and the operations it supports.

Variants

Uniform(UniformDistribution)

Exponential(ExponentialDistribution)

Gaussian(GaussianDistribution)

Bernoulli(BernoulliDistribution)

Generic(GenericDistribution)

Implementations

impl Distribution

pub fn new_uniform(interval: Interval) -> Result<Distribution, DataFusionError>

Constructs a new Uniform distribution from the given Interval.

pub fn new_exponential( rate: ScalarValue, offset: ScalarValue, positive_tail: bool, ) -> Result<Distribution, DataFusionError>

Constructs a new Exponential distribution from the given rate/offset pair, and validates the given parameters.

pub fn new_gaussian( mean: ScalarValue, variance: ScalarValue, ) -> Result<Distribution, DataFusionError>

Constructs a new Gaussian distribution from the given mean/variance pair, and validates the given parameters.

pub fn new_bernoulli(p: ScalarValue) -> Result<Distribution, DataFusionError>

Constructs a new Bernoulli distribution from the given success probability, and validates the given parameters.

pub fn new_generic( mean: ScalarValue, median: ScalarValue, variance: ScalarValue, range: Interval, ) -> Result<Distribution, DataFusionError>

Constructs a new Generic distribution from the given mean, median, variance, and range values after validating the given parameters.

pub fn new_from_interval( range: Interval, ) -> Result<Distribution, DataFusionError>

Constructs a new Generic distribution from the given range. Other parameters (mean, median and variance) are initialized with null values.

pub fn mean(&self) -> Result<ScalarValue, DataFusionError>

Extracts the mean value of this uncertain quantity, depending on its distribution:

• A Uniform distribution’s interval determines its mean value, which is the arithmetic average of the interval endpoints.
• An Exponential distribution’s mean is calculable by the formula offset + 1 / λ, where λ is the (non-negative) rate.
• A Gaussian distribution contains the mean explicitly.
• A Bernoulli distribution’s mean is equal to its success probability p.
• A Generic distribution may have it explicitly, or this information may be absent.

pub fn median(&self) -> Result<ScalarValue, DataFusionError>

Extracts the median value of this uncertain quantity, depending on its distribution:

• A Uniform distribution’s interval determines its median value, which is the arithmetic average of the interval endpoints.
• An Exponential distribution’s median is calculable by the formula offset + ln(2) / λ, where λ is the (non-negative) rate.
• A Gaussian distribution’s median is equal to its mean, which is specified explicitly.
• A Bernoulli distribution’s median is 1 if p > 0.5 and 0 otherwise, where p is the success probability.
• A Generic distribution may have it explicitly, or this information may be absent.

pub fn variance(&self) -> Result<ScalarValue, DataFusionError>

Extracts the variance value of this uncertain quantity, depending on its distribution:

• A Uniform distribution’s interval determines its variance value, which is calculable by the formula (upper - lower) ^ 2 / 12.
• An Exponential distribution’s variance is calculable by the formula 1 / (λ ^ 2), where λ is the (non-negative) rate.
• A Gaussian distribution’s variance is specified explicitly.
• A Bernoulli distribution’s median is given by the formula p * (1 - p) where p is the success probability.
• A Generic distribution may have it explicitly, or this information may be absent.

pub fn range(&self) -> Result<Interval, DataFusionError>

Extracts the range of this uncertain quantity, depending on its distribution:

• A Uniform distribution’s range is simply its interval.
• An Exponential distribution’s range is [offset, +∞).
• A Gaussian distribution’s range is unbounded.
• A Bernoulli distribution’s range is Interval::UNCERTAIN, if p is neither 0 nor 1. Otherwise, it is Interval::CERTAINLY_FALSE and Interval::CERTAINLY_TRUE, respectively.
• A Generic distribution is unbounded by default, but more information may be present.

pub fn data_type(&self) -> DataType

Returns the data type of the statistical parameters comprising this distribution.

pub fn target_type(args: &[&ScalarValue]) -> Result<DataType, DataFusionError>

Trait Implementations

impl Clone for Distribution

fn clone(&self) -> Distribution

Returns a copy of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for Distribution

fn fmt(&self, f: &mut Formatter<'_>) -> Result<(), Error>

Formats the value using the given formatter.

impl PartialEq for Distribution

fn eq(&self, other: &Distribution) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for Distribution