This is so interesting in the main building blocks of our dataset. You are called to ser features which can be categorical or numerical. But to work with algorithmic trading you need an actual draft working. Simply put, categorical data into the group data with similar features while numerical data gives information with numbers of trading outcomes.