I was trying to train a randomforest classifier in python. However, in my original pandas.dataframe, there are float64, object, datetime64, int64 and bool dtypes(nearly all kinds of dtypes allowed in pandas).
- Is it necessary to convert a bool to float or int?
- For a two-value object column, should I convert it to bool, int, or float? Which one would perform better? Or does it not matter?
Thanks!