To all modelers, can anyone explain why when I use the retrospective approach to generate my MACC, I encounter calculation errors that result in negative percentage share values for technology? However, when I check the model scenarios, there are no negative technological shares, and when I use the partial approach, my MACC is successfully generated. Can someone explain the numerical method used in generating MACC retrospectively? It seems like the calculation is performed by overriding one scenario with another, causing the technology share values for my "clean cooking" scenario to become negative.