reset

Reset incremental dynamic k-means clustering model

Since R2025a

Syntax

Mdl = reset(Mdl)

Description

Mdl = reset(Mdl) returns the incrementalDynamicKMeans model Mdl with reset k-means clustering properties. The function resets these properties:

• IsWarm to false

• NumClusters and NumDynamicClusters to their initial values after object creation

• Centroids and DynamicCentroids to their initial sizes after object creation, and their values to NaN

• ClusterCounts and DynamicClusterCounts to their initial sizes after object creation, and their values to 0

• NumTrainingObservations to 0

• Metrics and DynamicMetrics to NaN

reset preserves the NumPredictors, NumAdditionalClusters, EstimationPeriod, and WarmupPeriod properties of Mdl. However, if WarmupPeriod is 0, the reset function resets WarmupPeriod to the default value of 1000.

Examples

Reset Incremental Dynamic k-Means Clustering Model

Load the ionosphere data set.

load ionosphere.mat

Create an incremental dynamic k-means clustering model object Mdl with two clusters, and fit the model to the ionosphere data.

Mdl = incrementalDynamicKMeans(NumClusters=2);
Mdl = fit(Mdl,X);

Display the property values of the fitted model.

details(Mdl)

  incrementalDynamicKMeans with properties:

      NumAdditionalClusters: 10
              MergeClusters: 0
             MaxNumClusters: Inf
        GrowthPenaltyFactor: 10
           DynamicCentroids: [13×34 double]
       DynamicClusterCounts: [13×1 double]
         NumDynamicClusters: 13
             DynamicMetrics: [1×2 table]
                         Mu: []
                      Sigma: []
           EstimationPeriod: 0
                  Centroids: [13×34 double]
              ClusterCounts: [13×1 double]
                   Distance: "sqeuclidean"
           ForgettingFactor: 0.0500
                NumClusters: 13
                     IsWarm: 0
    NumTrainingObservations: 351
              NumPredictors: 34
               WarmupPeriod: 1000
                    Metrics: [1×2 table]
          MetricsWindowSize: 200

  Methods, Superclasses

Reset the model by calling the reset function.

Mdl = reset(Mdl);

Display the properties of the reset model.

details(Mdl)

  incrementalDynamicKMeans with properties:

      NumAdditionalClusters: 10
              MergeClusters: 0
             MaxNumClusters: Inf
        GrowthPenaltyFactor: 10
           DynamicCentroids: [11×34 double]
       DynamicClusterCounts: [11×1 double]
         NumDynamicClusters: 11
             DynamicMetrics: [1×2 table]
                         Mu: []
                      Sigma: []
           EstimationPeriod: 0
                  Centroids: [11×34 double]
              ClusterCounts: [11×1 double]
                   Distance: "sqeuclidean"
           ForgettingFactor: 0.0500
                NumClusters: 11
                     IsWarm: 0
    NumTrainingObservations: 0
              NumPredictors: 34
               WarmupPeriod: 1000
                    Metrics: [1×2 table]
          MetricsWindowSize: 200

  Methods, Superclasses

The reset function resets the following properties of Mdl: IsWarm, NumClusters, NumDynamicClusters, Centroids, DynamicCentroids, ClusterCounts, DynamicClusterCounts, Metrics, DynamicMetrics, and NumTrainingObserations.

Input Arguments

Mdl — Incremental dynamic k-means clustering model

incrementalDynamicKMeans model object

Incremental dynamic k-means clustering model, specified as an incrementalDynamicKMeans model object. You can create Mdl by calling incrementalDynamicKMeans directly.

Version History

Introduced in R2025a