Clustering with kmeans method and hierarchy clustering with agnes method.

basicClustering.R

library(dplyr)
library(cluster)
library(stringr)
library(caret)

set.seed(2342)

# Data preprocessing

raw_data <- 
  read.csv('lab7_data.csv', stringsAsFactors = F) %>% 
    mutate(Class = str_replace_all(Class, " ", "")) %>% 
    mutate(Class = factor(Class))

unclassified_data <-
  raw_data %>% 
    select(-Class)

# Computations

kmeans_cluster <- kmeans(unclassified_data, 3)

clusterisation_error <- 
  confusionMatrix(kmeans_cluster$cluster, as.numeric(raw_data$Class)) %>%
  (function (matrix) 1 - matrix$overall[['Accuracy']])
  
plot(unclassified_data$Sepala.length,
     col = kmeans_cluster$cluster,
     ylab = "Sepala length",
     xlab = "Observation",
     main = paste0("kmeans clusterization with error=", clusterisation_error))

hierarchy_cluster <- agnes(unclassified_data)

plot(hierarchy_cluster,
     xlab = "cluster",
     ylab = "height",
     main = "hierarchy clusterization")