k-means/ Clarans - how to deal with categorical data and how to choose the number of clusters

Question

I am using a dataset that contains both numerical and categorical data.

Let\'s say I have a column called categorical_data that can have multiple values