Accuracy score of a Decision Tree Classifier

Question

import sys
from class_vis import prettyPicture
from prep_terrain_data import makeTerrainData
from sklearn.tree import DecisionTreeClassifier
from sklearn.metrics import accuracy_score

import numpy as np
import pylab as pl

features_train, labels_train, features_test, labels_test = makeTerrainData()
X = features_train
Y = labels_train
clf = DecisionTreeClassifier()
clf = clf.fit(X,Y)
labels_test = clf.predict(features_test)

acc = accuracy_score(labels_test, labels_train)

I can't calculate the accuracy of a DecisionTreeClassifier using the above code. Can somebody help me with this?

Solution

The problem is that you are mixing up things. It doesn't mean anything to compute the accuracy comparing the train and test labels.

Do the following instead:

features_train, labels_train, features_test, labels_test = makeTerrainData()
X = features_train
Y = labels_train
clf = DecisionTreeClassifier()
clf = clf.fit(X,Y)
# Here call it somehing else!
yhat_test = clf.predict(features_test)
# Compute accuracy based on test samples
acc = accuracy_score(labels_test, yhat_test)

Comments

[Shridhar R Kulkarni]

Mentioning error here will help.

[MMF]

Well, it doesn't mean anything to compute the accuracy by comparing the labels of the test and the train, first, they are not related and second you most probably don't even have the same length for both! Your problem is that you overwrite the name labels_test, call it something else

[user269867]

global name 'accuracy_score' is not defined

[taga]

I get the error ValueError: labels_test and yhat_test have different number of output (2!=1)

[taga]

Is there any way that you can help me with this? stackoverflow.com/questions/56622349/…