On the documentation for XGBoost parameters it says:
colsample_by* parameters work cumulatively. For instance, the combination {'colsample_bytree':0.5, 'colsample_bylevel':0.5, 'colsample_bynode':0.5} with 64 features will leave 8 features to choose from at each split.
With this in mind, does the following formula make sense to determine the minimum possible colsample* values when attempting hyperparameter optimization?
(features * (x^3)) = 1