fe_cluster: Feature Engineering Clustering

Description Usage Arguments Value

View source: R/fe_cluster.R

Description

This function takes in a data table and a desired number of features in a cluster. This function will filter only numerical features first and cluster them with randomly picked numerical features. the final data table will have added clustered id's and distance to centroids added

Usage

1
fe_cluster(DT, test, predInClust = 5)

Arguments

a

train data table

a

test data table

number

of k in kmeans clustering

Value

a list with finished two data tables


tw1118/myutilitypackage documentation built on May 16, 2019, 9:13 p.m.