rxor: Random XOR Samples
In gencve: General Cross Validation Engine

Description Usage Arguments Details Value Author(s) Examples

Data is generated for the XOR problem. The default settings produce a data.frame with columns x1, x2 and y and with 4 rows and this table defines the XOR problem. The output y is defined by the XOR operation applied to the Boolean x1 and x2.

1	rxor(n = 1, p = 0)

`n`	sample size is 4*n
`p`	extra random inputs, x3, x4, ... etc. So the output data frame has dimensions 4*n by 2+p+1 columns. The extra p columns are random Bernouilli random variables with equi-probable outcomes.

This was a famous problem in online learning.

data.frame with 4*n rows and 2+p+1 columns. The last column corresponds to the output.

A. I. McLeod

1
2
3

library("C50")
Xy <- rxor(n=5, p=2)
C5.0(y ~ ., data=Xy)

Call:
C5.0.formula(formula = y ~ ., data = Xy)

Classification Tree
Number of samples: 20 
Number of predictors: 4 

Tree size: 6 

Non-standard options: attempt to group attributes