how can I program this log-likelyhood function?

[Jeongmin Yu]

Could anyone kindly help me program this log-likelyhood funtion in Stata?
Sorry inadvance, for the complicated fuction, but I'm so desperate to code this in Stata.

The function is as follows:

where f(δ) is normal distribution pdf function, and yij = 1 or 0.

I'd like to maximize above function (marked as "(4)") to estimate alpha(α), beta(βj), and delta(δij).

I'll really appreciate if anyone could help me.

Thank you.

[Phil Bromiley]

First, if this is a conventional estimator, then you'll find it infinitely easier to use Stata's routines rather than trying to program this yourself. Look at the logit, probit, and panel versions of such documentation.

Second, a lot of the summations can be done with egen by something.

But, if this needs to be programmed, it will need to be in the context of an optimizer.