How to calculate precision, recall and F-score with libSVM in python

Question

I want to calculate the precision, recall and f-score using libsvm in Python but I do not know how. I have found this site but I have not understand how to call the function, if you can help me through example.

Trull

Answer 1

You can take advantage of scikit-learn, which is one of the best packages for machine learning in Python. Its SVM implementation uses libsvmand you can work out precision, recall and f-score as shown in the following snippet:

from sklearn import svm
from sklearn import metrics
from sklearn.cross_validation import train_test_split
from sklearn.datasets import load_iris

# prepare dataset
iris = load_iris()
X = iris.data[:, :2]
y = iris.target
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

# svm classification
clf = svm.SVC(kernel='rbf', gamma=0.7, C = 1.0).fit(X_train, y_train)
y_predicted = clf.predict(X_test)

# performance
print "Classification report for %s" % clf
print
print metrics.classification_report(y_test, y_predicted)
print
print "Confusion matrix"
print metrics.confusion_matrix(y_test, y_predicted)

Which will produce an output similar to this:

Classification report for SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0, degree=3, gamma=0.7,
kernel=rbf, max_iter=-1, probability=False, shrinking=True, tol=0.001,
verbose=False)

             precision    recall  f1-score   support

          0       1.00      1.00      1.00         9
          1       0.90      0.69      0.78        13
          2       0.64      0.88      0.74         8

avg / total       0.86      0.83      0.84        30


Confusion matrix
[[9 0 0]
 [0 9 4]
 [0 1 7]]

Of course, you can use the libsvm tools you have mentioned, however they are designed to work only with binary classification whereas scikitallows you to work with multiclass.

Pinnatiped

What dataset you are loading ? For instance if my dataset is in the text file how to use them ? – Overstock

In the example I used a predefined dataset called iris that comes with scikit-learn. For a particular dataset you will need to convert your text data in numpy matrixes. – Pinnatiped