Struct MADDetector 

pub struct MADDetector { /* private fields */ }

A detector using the Median Absolute Deviation (MAD) to detect outliers.

Implementations

impl MADDetector

pub fn with_threshold(threshold: f64) -> MADDetector

Create a new MAD detector with the given threshold.

pub fn with_sensitivity(sensitivity: f64) -> Result<MADDetector, Error>

Create a new MAD detector with the given sensitivity.

At detection-time, a sensible value for threshold will be calculated using the scale of the data and the sensitivity value.

pub fn set_threshold(&mut self, threshold: f64)

Set threshold for the MAD detector.

Example

use augurs::outlier::MADDetector;

let mut detector = MADDetector::with_threshold(1.0);
detector.set_threshold(3.0);

pub fn set_sensitivity(&mut self, sensitivity: f64) -> Result<(), Error>

Set sensitivity for the MAD detector.

Example

use augurs::outlier::MADDetector;

let mut detector = MADDetector::with_sensitivity(0.5).expect("sensitivity is between 0.0 and 1.0");
detector.set_sensitivity(0.1).expect("sensitivity is between 0.0 and 1.0");

pub fn set_medians(&mut self, medians: Medians)

Set the precalculated medians.

The medians can be calculated using MADDetector::calculate_double_medians.

pub fn calculate_double_medians( data: &[&[f64]], ) -> Result<Medians, PreprocessingError>

Calculate the medians of the unprocessed data.

This can be used to precalculate the medians of a larger set of data, for example.

Errors

Returns an error if the data is empty, contains only NaNs, or if the lower or upper median is 0.0.

Trait Implementations

impl Clone for MADDetector

fn clone(&self) -> MADDetector

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for MADDetector

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

Formats the value using the given formatter.

impl OutlierDetector for MADDetector

type PreprocessedData = PreprocessedData

The preprocessed data used by the outlier detection algorithm.

fn preprocess( &self, y: &[&[f64]], ) -> Result<<MADDetector as OutlierDetector>::PreprocessedData, Error>

Preprocess the given slice of series.

fn detect( &self, y: &<MADDetector as OutlierDetector>::PreprocessedData, ) -> Result<OutlierOutput, Error>

Detect outliers in the given slice of series.