Re: [Gretl-devel] Forecast based on ordered probit
Am 16.12.2015 um 16:37 schrieb Artur Tarassow:
Dear all,
I just trying to run some forecasts based on an ordered probit
regression model and detected some (maybe?) slight inconsistency -- if any.
Using the fcast command returns for the BINARY PROBIT the estimated
probabilties, but for the ORDERED PROBIT a series of the predicted
response at each observation (the response for which the estimated
probability is highest).
I had nothing to do with the design, but it seems to make sense: For a
binary probit, giving the prob p_1 of the case 1 tells you everything,
because logically the other case must have 1-p_1. But with an ordered
probit you need at least two probs, and you cannot fit this information
into a single series.
Ok you could say for consistency the binary probit then also should give
just the predicted 0 and 1 after fcast.
Have you tried looking what's in the $fcast accessor? Maybe somebody
buried an easter egg in there for ordered probits.
good luck,
sven