clf.fit(X, y)
# 3. Predict.
if args.dataset == "iris":
input = [0, 0, 5.0, 1.5]
elif args.dataset == "wifi":
input = [-70, 0, 0, 0, -40, 0, 0]
elif args.dataset == "breast":
input = [np.random.rand() for _ in range(30)]
pred = clf.predict([input])[0]
print("Input: {}".format(input))
print("Prediction: " + dataset.target_names[pred])
# 4. Visualize.
if args.use_sklearn:
export_graphviz(
clf,
out_file="tree.dot",
feature_names=dataset.feature_names,
class_names=dataset.target_names,
rounded=True,
filled=True,
)
print("Done. To convert to PNG, run: dot -Tpng tree.dot -o tree.png")
else:
clf.debug(
list(dataset.feature_names),
list(dataset.target_names),
not args.hide_details,
)
