Apply glm to iris dataset in caret R

Question

I try to apply the glm algorithm on the Iris dataset, using the following code:

library(tidyverse)
library(caret)

dataset <- iris
tt_index <- createDataPartition(dataset$Sepal.Length, times = 1, p = 0.9, list = FALSE)
train_set <- dataset[tt_index, ]
test_set <- dataset[-tt_index, ]

model_glm <- train(Species ~., 
                   data = train_set,
                   method = "glm")

But it returned me this alert:

Something is wrong; all the Accuracy metric values are missing:
    Accuracy       Kappa    
 Min.   : NA   Min.   : NA  
 1st Qu.: NA   1st Qu.: NA  
 Median : NA   Median : NA  
 Mean   :NaN   Mean   :NaN  
 3rd Qu.: NA   3rd Qu.: NA  
 Max.   : NA   Max.   : NA  
 NA's   :1     NA's   :1    

Perhaps I am missing something, please your help will be greatly appreciated.

Answer 1

You're trying to train a binary (ie binomial) classification model to data with a response variable that has more than 2 levels. The warnings that you are getting and that you can see if you type warnings() will tell you that

glm models can only use 2-class outcomes

So this won't work.

An option is to omit one of the outcomes, eg do

dataset <- subset(iris, Species != "virginica")
dataset <- transform(dataset, Species = droplevels(Species))

tt_index <- createDataPartition(
    dataset$Sepal.Length, times = 1, p = 0.5, list = FALSE)
train_set <- dataset[tt_index, ]
test_set <- dataset[-tt_index, ]

model_glm <- train(
    Species ~.,
    data = train_set,
    method = "glm",
    family = "binomial")

This will still give warnings, but they have a different origin. Bottom line is, that this is probably not a very good example for testing glm -based binomial classification.