createRandomSet: Create a set of historical data and test data.
In tkonopka/Rthresholdout: Re-usable holdout implementation in R

This function will output a list with four objects. Two objects will be data matrices representing historical and test data. Two objects will be signals or sample classes that describe the samples in both datasets.

1 2	createRandomSet(num.features, num.samples, nbiased = 20, bias = 6/sqrt(num.features))

`num.features`	- number of features in each dataset
`num.samples`	- number of samples in each dataset
`nbiased`	- number of features with a true correlation between feature value and signal value
`bias`	- a factor that determines the strength of correlations for the nbiased features