Write a program in either python or java to implement the K-means clustering algorithm. Your program should start by prompting the user to enter the file name where the data are stored and the number of clusters (K).

Assume that the dataset is stored as comma-separated values. The first line is a header for the attributes. The data records starts from the second line where every line represents one record. Your program should be able to ignore any attribute (based on the users input).

The results will show the centroids of the clusters, the number of instances in each class, the total square error, and the number iterations that was used to get this result.

You can use many available numerical datasets from the UCI Machine Learning Repository. You can find these datasets at: http://archive.ics.uci.edu/ml/datasets.html.