In the dataset there will be several numbers of attributes and classes of attributes. The measurement of homogeneity or heterogeneity of data in the dataset is based on classes. The purity of table can be identified by, which contains only one class. The data table which consist of more than several classes are known as heterogeneous or impurity of table. There are several ways to measure impurity of table impurity in the tables. But the well method is entropy, gini index and classification error.