plot.nb2: Plot Naive Bayes Models

In aschersleben/NBCD: Naive Bayes for Online Classification with Concept Drift

Description

Plot Naive Bayes Models

Usage

## S3 method for class 'nb2'
plot(x, xlim, ylim, features, sdevs = 2, gridsize, data,
  class.name = "class", ..., gg = TRUE, psize = 2.5)

Arguments

x	[nb2] A Naive Bayes Model
xlim, ylim	[numeric(2)]
features	[numeric(2) | character(2)] If missing, the first two features are plotted.
sdevs	[numeric(1)] If missing(xlim) & missing(ylim): How many standard deviations from the mean in x and y direction.
gridsize	[numeric(1)] Size of grid to be predicted and plotted.
data	[list | data.frame] List with elements "x" and "class" or data.frame (then see argument class.name).
class.name	[character] If data is data.frame, then give the class' column name here. Default: "class".
...	Currently ignored.
gg	[logical] Use ggplot2 for plotting? Highly recommended!
psize	[numeric] Point size for geom_point.

See Also

nb2, update.nb2, print.nb2, predict.nb2

Examples

mod <- nb2(iris[, 1:4], iris[, 5])
plot(mod, features = 2:3, xlim = c(1.5, 4.5), ylim = c(0.5, 7.5),
     gridsize = 75, gg = FALSE)
points(iris[, 2:3], col = iris[, 5], pch = 4, lwd = 3)

require(ggplot2)
mod1 <- nb2(iris[, 1:4], iris[, 5])
preds <- predict(mod1, iris[, c(1, 3)])
iris2 <- cbind(iris, err = (preds != iris$Species))
plot(mod1, features = c(1, 3), xlim = c(4, 8), ylim = c(0, 8)) +
  geom_point(data = subset(iris2, iris2$err),
             aes(x = Sepal.Length, y = Petal.Length, shape = Species),
             size = 3.5, col = "white", show.legend = FALSE) +
  geom_point(aes(x = Sepal.Length, y = Petal.Length, shape = Species),
             data = iris2, size = 2) +
  guides(colour = FALSE, shape = FALSE)

discParam <- list(Sepal.L = c(5, 6, 7), Sepal.W = c(2.5, 3.5))
mod2 <- nb2(iris[, 1:4], iris[, 5], discretize = "fixed",
            discParams = discParam)
preds <- predict(mod2, iris[, c(1, 3)])
iris2 <- cbind(iris, err = (preds != iris$Species))
plot(mod2, features = c(1, 3), xlim = c(4, 8), ylim = c(0, 8)) +
  geom_point(aes(x = Sepal.Length, y = Petal.Length, shape = Species),
             data = subset(iris2, iris2$err), size = 3.5, col = "white",
             show.legend = FALSE) +
  geom_point(aes(x = Sepal.Length, y = Petal.Length, shape = Species),
  data = iris2, size = 2) +
  guides(colour = FALSE, shape = FALSE)

aschersleben/NBCD documentation built on May 12, 2019, 4:32 a.m.