Random forests with scikit-learn: which variables decide the yield?

Afterwards you can train and tune a random forest with scikit-learn, rank its inputs by permutation importance, and plot their partial dependence.