pyrus-decision-tree

Decision Tree written in Rust, with Python bindings

Extremely slow Decision Tree written in Rust

Have you found yourself yerning for a slower version of Scikit-Learn's DecisionTreeClassifier with less features?

You've come to the right place.

This was a weekend project and I've botched the implementation somewhere, probably in the splitter logic for tree nodes, so while it's written entirely in Rust and requires no dependencies; it manages to do a lot of inefficient logic, very quickly.

On a handful of testing data, with equivelent parameters, it yields the same results, only much, much slower.

This is currently the first release and the tree only implements the scikit-learn API's fit(X, y) and predict(X) methods as of now, and only as a classifier (no regression tree yet)

Maybe I'll get back to fixing it, maybe not.

Install:

pip install --upgrade pyrus-decision-tree

Uninstall:

pip uninstall pyrus-decision-tree

Use:

from pyrus_decision_tree import PyrusDecisionTree

dataset = [[2.771244718, 1.7847839292],
           [1.728571309, 1.1697614132],
           [3.678319846, 2.812813571],
           [3.961043357, 2.619950321],
           [2.999208922, 2.209014212],
           [7.497545867, 3.162953546],
           [9.00220326,  3.339047188],
           [7.444542326, 0.476683375],
           [10.12493903, 3.234550982],
           [6.642287351, 3.319983761]]
targets = [0, 0, 0, 0, 0, 1, 1, 1, 1, 1]

clf = PyrusDecisionTree(5)
clf.fit(dataset, targets)
predictions = clf.predict(dataset)