Hi,
How categorical features are handled in sklearn ?
I have used R version which handles strings. Also AN6U5 says tree classifiers can handle categorical features (link - http://datascience.stackexchange.com/questions/5226/strings-as-features-in-decision-tree-random-forest)

But when I tried data which have city names and state names .. I can not pass them directly. I had to use LabelEncoder so that RandomForest in sklearn can handle. But now categorical data are numbers only .. how can RandomForest identify that certain column is categorical or not ?

Now my question is how it is handling numeric and categorical data at the same time ? or it is not handling ?