Features

[AyachiOmarChaima]

when using ecole.observation.NodeBipartite(), we will get the variable features (19-dimensions) and constraints features (5-dimensions), and one feature for edge
I want to know some features how did calculate such as the constraints features and the variables features (normed_reduced_cost, solution_value, solution_frac, scaled_age, average_incumbent_value, and the basis_status)
If someone can help me how did those features are calculated