mlbench (version 2.1-3.1)

mlbench.spirals: Two Spirals Benchmark Problem

Description

The inputs of the spirals problem are points on two entangled spirals. If sd>0, then Gaussian noise is added to each data point. mlbench.1spiral creates a single spiral.

Usage

mlbench.spirals(n, cycles=1, sd=0)
mlbench.1spiral(n, cycles=1, sd=0)

Value

Returns an object of class "mlbench.spirals" with components

x

input values

classes

factor vector of length n with target classes

Arguments

n

number of patterns to create

cycles

the number of cycles each spiral makes

sd

standard deviation of data points around the spirals

Examples

# 1 cycle each, no noise
p<-mlbench.spirals(300)
plot(p)
#
# 1.5 cycles each, with noise
p<-mlbench.spirals(300,1.5,0.05)
plot(p)