Auto K-Means with

K-Means Series

Steven P. Sanderson II, MPH




Many times in a project we want to perform some sort of clustering on a given set of data. This can be accomplished many different ways. This vignette will showcase how you can take a data set that is prepared, say like the internal iris file and process it with the function hai_kmeans_automl().

First lets take a look at the data itself.

df_tbl <- iris

#> Rows: 150
#> Columns: 5
#> $ Sepal.Length <dbl> 5.1, 4.9, 4.7, 4.6, 5.0, 5.4, 4.6, 5.0, 4.4, 4.9, 5.4, 4.…
#> $ Sepal.Width  <dbl> 3.5, 3.0, 3.2, 3.1, 3.6, 3.9, 3.4, 3.4, 2.9, 3.1, 3.7, 3.…
#> $ Petal.Length <dbl> 1.4, 1.4, 1.3, 1.5, 1.4, 1.7, 1.4, 1.5, 1.4, 1.5, 1.5, 1.…
#> $ Petal.Width  <dbl> 0.2, 0.2, 0.2, 0.2, 0.2, 0.4, 0.3, 0.2, 0.2, 0.1, 0.2, 0.…
#> $ Species      <fct> setosa, setosa, setosa, setosa, setosa, setosa, setosa, s…

From here we can see that the data is already prepared and ready to go. There is a factor column that denotes the species or the row data and the columns are already numeric. Now the rest is fairly simple and straight forward. Let’s use the hai_kmeans_automl() function to create the list output that comes from it where we will want to use the Species column as the predictor based upon the features presented.

Use the function

column_names <- names(iris)
target_col <- "Species"
predictor_cols <- setdiff(column_names, target_col)

Now we have our column inputs for the function, so we can go ahead and run it.


output <- hai_kmeans_automl(
  .data = df_tbl,
  .predictors = predictor_cols,
  .standardize = FALSE

h2o.shutdown(prompt = FALSE)

This function gives a lot of output inside of it. From here we will discuss what comes out of the function.

Function Output

Lets take a look at the structure of the output object. It is a list of lists with four main components. They are the following:

Lets explor each of these items.


Inside of the data list there are several sections. We can view and access these very simply. You will find that all of the outputs have been labeled in a very simple to understand manner.


Auto-ML Object

Now for the auto-ml object itself.


The Best Model

We also have in the output the best model that is saved off.


Scree Plot

There is also a ggplot2 scree plot that is generated, this helps us to understand how many clusters are in the data resulting from minimizing the within sum of squares errors.