Search code examples
rcluster-analysishierarchical-clusteringdendrogram

Can't Plot Dendrogram (invalid dendrogram input error)


I have 583 observations within my dataset, out of which I need to use 100 to make a Dendrogram. But I am getting the following error in the plot() function step.

how to troubleshoot/solve this?

##For 100 Observations
set.seed(44)
idx_100 <- sample(1:nrow(ilpd_df), 100)
distance_matrix_100 <- dist(as.matrix(ilpd_df[idx_100,-c(1,2,10,11)]), 
                        method = "euclidean") #Creates Hierarchical Clustering Solution

hc_100 <- hclust(distance_matrix_100)
plot(hc_100, hang = -1, labels=ilpd_df$Class) #CANT PLOT DUE TO INVALID INPUT ERROR

't plot


Solution

  • Could you put some minimal data that can be used to replicate the problem?

    I find no issue (R 3.6.1) with plot.hclust().

    The following simple code

    hc <- hclust(dist(USArrests)^2, "cen")
    class(hc)  # [1] "hclust"
    plot(hc, hang = -1, main = "Sample Tree", cex = .5)
    

    Produces

    Sample dendrogram