We thought we had this feature... but then it turns out that the unit tests were too shallow, didn't test that the sparse abundance matrix could participate correctly in the actual calculation. scipy.sparse implementations gum up the works in the power_mean functions.
We thought we had this feature... but then it turns out that the unit tests were too shallow, didn't test that the sparse abundance matrix could participate correctly in the actual calculation. scipy.sparse implementations gum up the works in the power_mean functions.