Open korenmiklos opened 1 year ago
Will not do this in the current report because it takes much deeper understanding than I thought. For some reason the groupby of pandas and the collapse of stata gives different number of observations. First n is 733, the second n is 745, while using the same variables that has the same number of unique observations. It seems that the combination of variables miss 12 observations in python compared to stata...
Run both on the same data. Asser dataframes are the same