I\'m working on a pandas dataframe in which certain features (columns) have hierarchical importance. Is there an algorithm that reduces dimensionality by disregarding some data,
