snp_tree: CART for Final SNP Selection
In jillbo1000/gwas3: Three Phase Filtering for GWAS

Creates classification and regression trees for final SNP selection using rpart. The function uses the dataset list from the elasticnet function in this package as input and returns a list of trees and the SNPs where the trees split. The trees are pruned using the 1-se rule.

1	snp_tree(data_list, y)

`data_list`	Object obtained from the `elasticnet` function.
`y`	A numeric vector of phenotypes for the data.

snp_tree returns a list for each dataset in data_list. Each of the sublists contains the pruned tree created by rpart, a vector containing the SNPs where the tree splits, and the alpha value associated with the elastic net model used to create the dataset. Each item is called using the list number and name, such as results[[1]]$tree and results[[1]]$snps.

`tree`	Tree produced by rpart. Tree is pruned using the 1-se rule.
`snps`	Vector of SNPs where the tree splits.
`alpha`	Alpha value from the elastic net model used to trim the input data.