How to obtain precision, recall and F-measure for clustering?

Precision and recall for a classification problem are obtained by the confusion matrix easily since the number of classes is provided in the problem. For a clustering problem, we do not have the number of classes, so some extra clusters may be presented in the solution.

Could anybody provide me with a Matlab code for the evaluation of results by precision, recall and F-measure in this case?

Thank you.

[Kartik Mittal]

Precision and recall are good performance measures for a supervised learning algorithm (where you have some ground truth). Since you are using some clustering algorithm, I am assuming you might not have any ground truth to validate your results. Correct me if I am wrong.