this analysis seems severely flawed - ranking row percentages? I mean it's a formatted table so we cant analyse it across the entire table rather on a row-by-row basis if we don't have the details. This challenge is way to extreme for microdata simulation so I am calling it a day