search

snoweye / EMCluster

EM Algorithm for Model-Based Clustering of Finite Mixture Gaussian Distribution
Mozilla Public License 2.0
17 stars 0 forks source link

Feature Request: $cluster and $centers should be added to the result #5

Open MovGP0 opened 6 years ago

MovGP0 commented 6 years ago

When using a clustering function like kmeans or dbscan::dbscan, the result data frame will have $cluster and $centers attributes. These attributes are currently missing from the result and should be added, such that the same functions can be used for analyzing and plotting the clustering result.

MovGP0 commented 6 years ago

correction: $cluster information is available. it is just called $class. Documentation is missing.

MovGP0 commented 6 years ago

correction: centers are also available. they are called $Mu. the row is the cluster ($Mu[1,] is the first cluster) and the columns is the position in the cluster (ie. $Mu[1,1] is the first measurement of the cluster).

The headers of $Mu should be renamed to the headers of the original data, such that it is easier to correlate them.