rfgb.tree module

Data structures and methods for learning decision trees.

class rfgb.tree.node(test=None, examples=None, information=None, level=None, parent=None, pos=None)

Bases: object

A node in a tree.

Parameters:

• expandQueue – Breadth first search node expansion strategy
• depth – initial depth is 0 because no node present
• maxDepth – max depth set to 1 because we want to at least learn a tree of depth 1
• learnedDecisionTree – this will hold all the clauses learned
• data – stores all the facts, positive and negative examples

data = None

depth = 0

expandOnBestTest(data=None)

Expand the node based on the best test.

expandQueue = []

getTrueExamples(clause, test, data)

Returns all examples that satisfy the clause with conjoined test literal.

static initTree(trainingData)

Create the root node of the tree.

static learnTree(data)

Method to create and learn the decision tree.

learnedDecisionTree = []

maxDepth = 1

static setMaxDepth(depth)

Set the maximum depth of the tree.