twidlr (version 0.0.0.9000)

kmeans: data.frame-first formula-second method for

Description

This function passes a data.frame, formula, and additional arguments to kmeans.

Usage

kmeans(data, formula = ~., ...)
# S3 method for kmeans
predict(object, data, ...)

Arguments

data

Data frame (or object that can be coerced to one) containing the variables in the model

formula

A formula (or object that can be coerced to one) with no response variable (optional). If omitted, all variables in "data" will be included.

...

Additional arguments to pass to model function

object

Fitted model

See Also

kmeans

Examples

# Standard kmeans
kmeans(iris, centers = 3)

# formula interface can be used to select certain variables
kmeans(iris, ~ Petal.Width + Sepal.Width, centers = 3)

#... or create new variables. Consider non-linear example:

# Create data for two circles
x <- matrix(rnorm(300),nc=2)
y <- x/sqrt(rowSums(x^2))
d <- data.frame(rbind(y, y*5))
plot(d)

fit <- kmeans(d, centers = 2)
plot(d, col = predict(fit, d))

fit <- kmeans(d, ~ X1 + X2 + I(X1^2 + X2^2), centers = 2)
plot(d, col = predict(fit, d))