outlierImputerComponent

Pipeline component for imputing outlier values

Since R2026a

Description

outlierImputerComponent is a pipeline component that performs outlier imputation. The pipeline component uses the functionality of the filloutliers function during the learn phase to identify and impute outlier values for a set of observations. During the run phase, the component uses the values learned during the learn phase to impute outlier values in a new data set.

Creation

Syntax

component = outlierImputerComponent

component = outlierImputerComponent(Name=Value)

Description

component = outlierImputerComponent creates a pipeline component for imputing outlier values.

component = outlierImputerComponent(Name=Value) sets writable Properties using one or more name-value arguments. For example, you can specify the outlier detection method by using the FindMethod name-value argument.

Properties

Learn Parameters

The software sets learn parameters when you create the component. You can modify learn parameters using dot notation any time before you use the learn object function. Any unset learn parameters use the corresponding default values.

FindMethod — Outlier detection method

"median" (default) | "gesd" | "grubbs" | "mean" | "percentiles" | "quartiles"

Outlier detection method, specified as one of the following values.

Value	Description
"gesd"	For each variable, find outliers by using the generalized extreme Studentized deviate test for outliers. Use ThresholdFactor to specify the alpha value for the test.
"grubbs"	For each variable, find outliers by using Grubbs’ test, which removes one outlier per iteration based on hypothesis testing. Use ThresholdFactor to specify the alpha value for the test.
"mean"	For each variable, outliers are values more than a certain number of standard deviations from the mean. Use ThresholdFactor to specify the number of standard deviations.
"median"	For each variable, outliers are values more than a certain number of scaled median absolute deviations (MAD) from the median. Use ThresholdFactor to specify the number of scaled MAD.
"percentiles"	For each variable, outliers are values below the lower threshold or above the upper threshold, as specified by Threshold.
"quartiles"	For each variable, outliers are values more than a certain number of interquartile ranges below the lower quartile (25 percent) or above the upper quartile (75 percent). Use ThresholdFactor to specify the number of interquartile ranges.

For more information, see findmethod.

Example: c = outlierImputerComponent(FindMethod="mean")

Example: c.FindMethod = "quartiles"

Data Types: char | string

ImputationMethod — Outlier imputation method

"center" (default) | "clip"

Outlier imputation method, specified as "center" or "clip".

• If you specify "center", the software determines the center value using the outlier detection method (FindMethod).

• If you specify "clip", the software determines the lower and upper thresholds using the outlier detection method (FindMethod).

Example: c = outlierImputerComponent(ImputationMethod="clip")

Example: c.ImputationMethod = "center"

Data Types: char | string

ThresholdFactor — Outlier detection threshold factor

nonnegative scalar

Outlier detection threshold factor, specified as a nonnegative scalar.

• When FindMethod is "median", the outlier detection threshold factor is the number of scaled MAD, which is 3 by default.

• When FindMethod is "mean", the outlier detection threshold factor is the number of standard deviations from the mean, which is 3 by default.

• When FindMethod is "grubbs" or "gesd", the outlier detection threshold factor is a scalar in the interval (0,1), which represents the alpha value of the hypothesis test. Values close to 0 result in a smaller number of outliers, and values close to 1 result in a larger number of outliers. The default value is 0.05.

• When FindMethod is "quartiles", the outlier detection threshold factor is the number of interquartile ranges, which is 1.5 by default.

You cannot specify ThresholdFactor when the outlier detection method is "percentiles".

Example: c = outlierImputerComponent(ThresholdFactor=2.5)

Example: c.ThresholdFactor = 0.01

Data Types: single | double

Threshold — Lower and upper percentile thresholds

nonnegative vector with two elements

Lower and upper percentile thresholds, specified as a nonnegative vector with two elements in the interval [0,100]. The first element indicates the lower percentile threshold, and the second element indicates the upper percentile threshold. The first element must be less than the second element.

You must specify Threshold when the outlier detection method (FindMethod) is "percentiles". You cannot specify Threshold for any other outlier detection method.

Example: c = outlierImputerComponent(Threshold=[10 90])

Example: c.Threshold=[5 95]

Data Types: single | double

MaxNumOutliers — Maximum number of outliers to impute

positive integer scalar

Maximum number of outliers to impute, specified as a positive integer scalar.

If you do not specify the MaxNumOutliers value, the software uses the integer nearest to 10 percent of n, where n is the number of observations in the data argument of learn.

You can specify MaxNumOutliers only when the outlier detection method (FindMethod) is "gesd".

Example: c = outlierImputerComponent(MaxNumOutliers=20)

Example: c.MaxNumOutliers = 5

Data Types: single | double

Component Properties

The software sets component properties when you create the component. You can modify the component properties (excluding HasLearnables and HasLearned) using dot notation at any time. You cannot modify the HasLearnables and HasLearned properties directly.

Name — Component identifier

"OutlierImputer" (default) | character vector | string scalar

Component identifier, specified as a character vector or string scalar.

Example: c = outlierImputerComponent(Name="OutlierImputation")

Example: c.Name = "Imputation"

Data Types: char | string

Inputs — Names of input ports

"DataIn" (default) | character vector | string array | cell array of character vectors

Names of the input ports, specified as a character vector, string array, or cell array of character vectors.

Example: c = outlierImputerComponent(Inputs="X")

Example: c.Inputs = "X1"

Data Types: char | string | cell

Outputs — Names of output ports

["DataOut","IsOutlier"] (default) | character vector | string array | cell array of character vectors

Names of the output ports, specified as a character vector, string array, or cell array of character vectors.

Example: c = outlierImputerComponent(Outputs=["newX","indices"])

Example: c.Outputs = ["X1","Idx"]

Data Types: char | string | cell

InputTags — Tags that enable automatic connection of component inputs

1 (default) | nonnegative integer vector

Tags that enable the automatic connection of the component inputs with other components or pipelines, specified as a nonnegative integer vector. If you specify InputTags, the number of tags must match the number of inputs in Inputs.

Example: c = outlierImputerComponent(InputTags=0)

Example: c.InputTags = 1

Data Types: single | double

OutputTags — Tags that enable automatic connection of component outputs

[1 0] (default) | nonnegative integer vector

Tags that enable the automatic connection of the component outputs with other components or pipelines, specified as a nonnegative integer vector. If you specify OutputTags, the number of tags must match the number of outputs in Outputs.

Example: c = outlierImputerComponent(OutputTags=[0 0])

Example: c.OutputTags = [1 0]

Data Types: single | double

HasLearnables — Indicator for learnables

Read-only: 1 (true) (default)

This property is read-only.

Indicator for the learnables, returned as 1 (true). A value of 1 indicates that the component contains Learnables.

Data Types: logical

HasLearned — Indicator showing learning status of component

Read-only: 0 (false) (default) | 1 (true)

This property is read-only.

Indicator showing the learning status of the component, returned as 0 (false) or 1 (true). A value of 1 indicates that the learn object function has been applied to the component, and the Learnables are nonempty.

Data Types: logical

Learnables

The software sets learnables when you use the learn object function. You cannot modify learnables directly.

LowerThreshold — Lower threshold for identifying outliers

Read-only: table | []

This property is read-only.

Lower threshold for identifying outliers, returned as a table. Each value corresponds to a variable in VariablesWithOutliers.

UpperThreshold — Upper threshold for identifying outliers

Read-only: table | []

This property is read-only.

Upper threshold for identifying outliers, returned as a table. Each value corresponds to a variable in VariablesWithOutliers.

Center — Center value for imputing outliers

Read-only: table | []

This property is read-only.

Center value for imputing outliers, returned as a table. Each value corresponds to a variable in VariableWithOutliers.

VariablesWithOutliers — Names of variables used by component

Read-only: string array | []

This property is read-only.

Names of the variables used by the component to derive the LowerThreshold, UpperThreshold, and Center values, returned as a string array. The variables correspond to columns in the data argument of learn.

Object Functions

learn	Initialize and evaluate pipeline or component
run	Execute pipeline or component for inference after learning
reset	Reset pipeline or component
series	Connect components in series to create pipeline
parallel	Connect components or pipelines in parallel to create pipeline
view	View diagram of pipeline inputs, outputs, components, and connections

Examples

Create Component for Imputing Outlier Values

Create a pipeline component that imputes outlier values in observations.

component = outlierImputerComponent

component = 

  outlierImputerComponent with properties:

                     Name: "OutlierImputer"
                   Inputs: "DataIn"
                InputTags: 1
                  Outputs: ["DataOut"    "IsOutlier"]
               OutputTags: [1 0]

   
Learnables (HasLearned = false)
           LowerThreshold: []
           UpperThreshold: []
                   Center: []
    VariablesWithOutliers: []


Show all parameters

component is an outlierImputerComponent object that contains four learnables: LowerThreshold, UpperThreshold, Center, and VariablesWithOutliers. The properties remain empty until you pass data to the component during the learn phase.

Load the carbig data set. Create a table containing the variables Acceleration, Displacement, and Horsepower.

load carbig
cars = table(Acceleration,Displacement,Horsepower);

Use the learn object function to find and impute outlier values in cars.

[component,newcars] = learn(component,cars);
component

component = 

  outlierImputerComponent with properties:

                     Name: "OutlierImputer"
                   Inputs: "DataIn"
                InputTags: 1
                  Outputs: ["DataOut"    "IsOutlier"]
               OutputTags: [1 0]

   
Learnables (HasLearned = true)
           LowerThreshold: [1×3 table]
           UpperThreshold: [1×3 table]
                   Center: [1×3 table]
    VariablesWithOutliers: ["Acceleration"    "Displacement"    "Horsepower"]


Show all parameters

The LowerThreshold, UpperThreshold, Center, and VariablesWithOutliers properties are nonempty, and the HasLearned property is set to true.

The newcars data set contains the same observations as cars, but has imputed outlier values. For example, the sixth observation in cars, which has outlier values for Displacement and Horsepower, uses imputed values in newcars.

observation = cars(6,:)
newObservation = newcars(6,:)

observation =

  1×3 table

    Acceleration    Displacement    Horsepower
    ____________    ____________    __________

         10             429            198    


newObservation =

  1×3 table

    Acceleration    Displacement    Horsepower
    ____________    ____________    __________

         10             151             95    

Version History

Introduced in R2026a