The code given clusters and selects features based on the calculated correlation between the selected columns, fix the code so that the calcualted dissimilarity matrix is symmetric, so it can be used to calculate Z and the labels. 