Open kenny5033 opened 1 month ago
See #75 for a possible approach to solving.
But note: If determining the columns can be done algorithmically, it may be better to just implement the algorithm rather than to maintain a hand-curated list. That way things would continue to work without additional effort if new categories are added later.
Idea: Have a column in the Index table containing which column in the diagram a node should go into. The current algorithm would be run, but if there is a value in this new column (column would be nullable) then that should override the default algorithm.
Possible other library
https://observablehq.com/@mariodelgadosr/sankey-diagram-with-draggable-nodes
Idea: Have parameters that let a user choose where they want the energy carriers to go. Options could include "all in column 4", "all in column 2", "default".
We have to decide where certain carriers should be placed on the sankey diagrams. Right now, the problem is that some energy carrier nodes should up in columns that cause backwards flows. This looks strange and could possible give users the wrong impression of energy flowing at first glance.
Instead the energy flows need to be place in columns that minimize backwards flows.
Ideas: