Struct PartialCorrelationTest 

pub struct PartialCorrelationTest {
    pub alpha: f64,
}

Partial correlation test using Fisher’s z-transform.

Under the null hypothesis of conditional independence (X independent Y | Z), the Fisher-transformed partial correlation is approximately N(0,1) for Gaussian data with sufficiently large n.

The partial correlation is computed via recursive formula or OLS residuals.

Fields

alpha: f64

Significance level (default 0.05).

Implementations

impl PartialCorrelationTest

pub fn new(alpha: f64) -> Self

Create a new test with the given significance level.

pub fn partial_correlation( &self, x: usize, y: usize, z_set: &[usize], data: ArrayView2<'_, f64>, ) -> StatsResult<f64>

Compute the partial correlation between x and y given z_set.

Trait Implementations

impl Clone for PartialCorrelationTest

fn clone(&self) -> PartialCorrelationTest

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl ConditionalIndependenceTest for PartialCorrelationTest

fn test( &self, x: usize, y: usize, z_set: &[usize], data: ArrayView2<'_, f64>, ) -> StatsResult<CITestResult>

Test whether variable x is independent of variable y given z_set, using column indices into data (rows = observations, cols = variables).

fn is_independent( &self, x: usize, y: usize, z_set: &[usize], data: ArrayView2<'_, f64>, alpha: f64, ) -> StatsResult<bool>

Convenience: test and return true if independent at level alpha.

impl Debug for PartialCorrelationTest

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

Formats the value using the given formatter.

impl Default for PartialCorrelationTest

fn default() -> Self

Returns the “default value” for a type.