Dear all,
I am running a panel dataset mixed logit with unlabelled alternatives.
To simplify, in my case the alternatives are tennis tournaments with different characteristics (for instance, prize money). The players' charactertistics ( for instance, ranking) interact with the tournaments' characteristics as they do not vary across alternatives.
My question is: in a postestimation phase, how can I calculate the probability to chose a tournament with, for instance, a 1000$ higher prize money as a function of the ranking? I know that margins doesn't work with unlabelled alternatives.
Thank you
I am running a panel dataset mixed logit with unlabelled alternatives.
To simplify, in my case the alternatives are tennis tournaments with different characteristics (for instance, prize money). The players' charactertistics ( for instance, ranking) interact with the tournaments' characteristics as they do not vary across alternatives.
My question is: in a postestimation phase, how can I calculate the probability to chose a tournament with, for instance, a 1000$ higher prize money as a function of the ranking? I know that margins doesn't work with unlabelled alternatives.
Thank you
