Implements the Fratar Factoring algorithm (aka Biproportional Matrix Factoring).
Still TODO:
Need to get origin/destination targets from the user. The code currently uses placeholder targets.
The margin stats on the difference matrix (model.od_diff) do not make sense and are not intuitive (sum, target, delta). Should these margin stats be removed?
Implements the Fratar Factoring algorithm (aka Biproportional Matrix Factoring).
Still TODO:
model.od_diff
) do not make sense and are not intuitive (sum, target, delta). Should these margin stats be removed?