visualise: Produce a visualisation of a k-nearest neighbours.

In phil8192/lazy-iris: Lazy Learning on the Iris Dataset

Description

Given a query, it's resulting neighbours and the original feature space, this function will plot a visualisation of the position of the query in relation to it's neighbours in the feature space.

Usage

visualise(feature.space, class.name, query = NULL, neighbours = NULL,
  plot.hist = F, plot.cor = F, ...)

Arguments

feature.space	The feature space.
class.name	The class name. E.g., species.
query	The query (optional).
neighbours	The resulting k-nearest neighbours (optional).
plot.hist	If TRUE plot a feature distribution histogram.
plot.cor	If TRUE plot feature correlation.
...	Additional arguments to pairs plotting function.

Details

Colour code:

• rediris setosa

• greeniris versicolour

• blueiris virginica

• blackquery

The visualisation makes use of R's pairs plotting function, more information on which can be viewed in the function's man page: help(pairs).

Author(s)

phil

Examples

# plot all the data.
visualise(iris.data, class.name="species", main="iris data", plot.hist=TRUE,
    plot.cor=TRUE)

# do not plot the first 2 features and omit the histogram+correlation plots.
visualise(iris.data[, -(1:2)], class.name="species", main="iris data")

#### visualise k-nearest neighbours.

# form a query.
q <- list(
    sepal.length=5.84,
    sepal.width=3.05,
    petal.length=3.76,
    petal.width=1.20)

# get the 10-nearest neighbours
top.10 <- knn(q, iris.data, 10)

# visualise the neighbours and query point.
visualise(iris.data, class.name="species", query=q, neighbours=top.10,
    main="iris data neighbours", plot.hist=TRUE, plot.cor=FALSE)

phil8192/lazy-iris documentation built on May 25, 2019, 2:56 a.m.