Hi Statalist!
I have a dataset that contains 3 variables (id, math and reading) and 20 observations.
I would like to create 10 pairs so as to minimize the sum of Mahalanobis distances between math and reading within pairs. (I will use this to decide whom to treat in an experiment).
I have tried using the mahaselectunique command. However, this command requires a variable that specifies who is "treated" so that it can assign each untreated to a treated variable. Furthermore, I have not even succeeded in using mahaselectunique correctly (I am a Stata newbie).
I would be very grateful if someone could show me how to get a new variable that specifies the 10 "optimized pairs" I need.
I have a dataset that contains 3 variables (id, math and reading) and 20 observations.
I would like to create 10 pairs so as to minimize the sum of Mahalanobis distances between math and reading within pairs. (I will use this to decide whom to treat in an experiment).
I have tried using the mahaselectunique command. However, this command requires a variable that specifies who is "treated" so that it can assign each untreated to a treated variable. Furthermore, I have not even succeeded in using mahaselectunique correctly (I am a Stata newbie).
I would be very grateful if someone could show me how to get a new variable that specifies the 10 "optimized pairs" I need.
